{"id":"RAN-DBDEA-00001","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:95e2004762bfb8f09736098cd277999e99d68f0592670db5630934343a8f4233","aliases":[],"snapshot":{"statement":"A data engineering team runs a nightly pipeline that reads Delta tables, applies PySpark transformations, and writes a gold table. The pipeline is executed by a Lakeflow Job and must use dedicated compute that starts when the job starts and is terminated automatically when the job finishes. Analysts must not be able to attach to this compute for interactive sessions.\n\nWhich compute option should the team use for this workload?","options":["An all-purpose cluster that stays running so analysts can attach to it between job runs","A job cluster scoped to the Lakeflow Job, created and terminated by the job itself","A serverless SQL warehouse configured for the notebook tasks in the job","A high-concurrency cluster shared with other teams that run ad hoc queries"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/compute","referenceTitle":"Compute for Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00002","exam":"DBDEA","topics":["platform"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:9dbabe31eaadd28b71703f96db521058070ae50632cbe4b7c05157a2ca9a52bc","aliases":[],"snapshot":{"statement":"A company currently exports data from its operational systems to Parquet files in cloud object storage. Analysts want to run ad hoc SQL queries over that data, and data scientists want to train models on the exact same tables. Both teams need ACID guarantees, reliable schema, and a single source of truth — without maintaining two separate copies of the data in a warehouse and a data lake.\n\nWhich approach directly addresses these requirements on the Databricks Data Intelligence Platform?","options":["Keep two copies of the data: load it into a classic data warehouse for the analysts and keep the raw files in the lake for the data scientists","Adopt the lakehouse approach: use the Delta Lake open table format on cloud object storage, with Unity Catalog governing access for all teams","Store everything as ephemeral in-memory DataFrames and recompute them from source whenever either team needs the data","Convert all data to unstructured JSON files so both teams can parse them with their own custom readers"],"answer":"B","reference":"https://docs.databricks.com/aws/en/introduction","referenceTitle":"Introduction to the Databricks Data Intelligence Platform"}}
{"id":"RAN-DBDEA-00003","exam":"DBDEA","topics":["platform"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:032643847bfa76f2dcf3bddd1570476c9538b4bbc779194a0cdad5feb7efddb5","aliases":[],"snapshot":{"statement":"A healthcare company is migrating its analytics to the Databricks Data Intelligence Platform. Its security officers need one central place to define who can access which tables across all workspaces, to track lineage, and to audit query activity for compliance reporting. They do not want to manage access rules separately in each cloud storage account.\n\nWhich platform component is the foundation for these requirements?","options":["Unity Catalog, the lakehouse's unified governance layer, with centralized permissions, lineage, and auditing","Delta Lake, which enforces access rules inside the table format through the ACID transaction log it maintains on every write","Serverless SQL warehouses, which automatically apply row-level filters to every query they execute","Lakeflow Jobs, which can grant and revoke privileges each time a scheduled task runs"],"answer":"A","reference":"https://docs.databricks.com/aws/en/introduction","referenceTitle":"Introduction to the Databricks Data Intelligence Platform"}}
{"id":"RAN-DBDEA-00004","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b326763b199ee056b4f06d5844ccf68cced322dcac6bba8dd87f3a11d7237671","aliases":[],"snapshot":{"statement":"A two-person data engineering team schedules a nightly Lakeflow Job that runs a PySpark notebook to refresh reporting tables. The team has no cluster administration expertise and wants to avoid selecting instance types, worker counts, or runtime versions. They also want the job's compute to become available quickly, with no infrastructure for them to deploy or configure.\n\nWhich compute option best fits these requirements?","options":["Serverless compute for jobs, which Databricks provisions, scales, and manages automatically","A classic job cluster with manually selected worker instance types and autoscaling bounds","A shared all-purpose cluster that remains running so the job can attach to it at night","A classic job cluster that the team resizes each evening based on the previous run"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/compute","referenceTitle":"Compute for Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00005","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6ad132e9cc01d24a4f67167baceb1802230b256a91699681af21add8587fc674","aliases":[],"snapshot":{"statement":"A BI team connects Tableau and Power BI to the Databricks workspace over JDBC/ODBC and spends the day running ad hoc SQL queries and dashboards against gold Delta tables. Analysts do not want to size or tune Spark clusters; they need compute that is optimized for SQL analytics and BI workloads and that starts and scales on demand with query traffic.\n\nWhich compute service should the team use?","options":["A classic all-purpose Spark cluster with Photon enabled, shared by all analysts","A SQL warehouse, the compute optimized for SQL queries, analytics, and business intelligence workloads","A classic job cluster scoped to a nightly Lakeflow Job, which the BI tools would have to connect to during the day while it stays running","Serverless compute for jobs, because the BI tools schedule their queries as job tasks"],"answer":"B","reference":"https://docs.databricks.com/aws/en/compute","referenceTitle":"Compute — Databricks Docs"}}
{"id":"RAN-DBDEA-00006","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:24a318ed2e0d3c8aaf5417480e0264519acbda62604a447f0ef05171664e83e5","aliases":[],"snapshot":{"statement":"A data engineering team is migrating a legacy Spark application into a Lakeflow Job. One task in the job must run a compiled JAR that the team maintains. The team would prefer to run the entire job on serverless compute for jobs to avoid managing clusters.\n\nWhich statement is correct?","options":["The JAR task can run on serverless compute for jobs once the JAR is uploaded as a dependent library","The JAR task must run on a serverless SQL warehouse, because SQL warehouses execute compiled artifacts","JAR tasks are not supported on serverless compute for jobs — that task needs classic jobs compute","Serverless compute for jobs supports every task type; the team only needs to pick a Databricks Runtime version"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/compute","referenceTitle":"Compute for Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00007","exam":"DBDEA","topics":["platform"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:dfab90fc6154888cb1151564d17d69a88b100a715b85856be761702f4305448a","aliases":[],"snapshot":{"statement":"A machine learning team works interactively in notebooks, training models on GPU-enabled instances with Databricks Runtime for Machine Learning and several custom environment libraries. To reduce administration, they ask whether this workload can move to serverless compute.\n\nWhat should the data engineer answer?","options":["Yes — serverless compute automatically attaches GPU instances when it detects ML libraries","Yes — serverless compute supports every Databricks Runtime version, including ML runtimes","No — interactive GPU training must be submitted as separate Lakeflow Jobs that each create a job cluster","No — ML Runtime on GPU is not supported on serverless compute; use a classic compute resource"],"answer":"D","reference":"https://docs.databricks.com/aws/en/compute/cluster-config-best-practices","referenceTitle":"Classic compute configuration best practices — Databricks Docs"}}
{"id":"RAN-DBDEA-00008","exam":"DBDEA","topics":["platform"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:5a047067fc317817f47b41a62f396ed837f3694cbf919f518b7916a473ede8d4","aliases":[],"snapshot":{"statement":"A retail company wants its data quality notebook to run automatically every night at 02:00, to retry the task if it fails, and to email the on-call engineer whenever a run fails. The notebook already exists in the workspace.\n\nWhich platform capability should the company use?","options":["Databricks Git Folders (formerly Repos), which run notebooks on a schedule from a Git branch","Unity Catalog, which governs data access and can also schedule notebook refreshes","Delta Lake time travel, which can rewind and replay a notebook's previous outputs","Lakeflow Jobs, which schedule the notebook with retries and failure notifications"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/","referenceTitle":"Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00009","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:64a1ef9f96cdb0a212ea212ce02523930dde701463345739cfb31160f7975ad1","aliases":[],"snapshot":{"statement":"A streaming media company is designing its ingestion layer on the Databricks Data Intelligence Platform. It has three sources:\n\n- A legacy billing system that delivers a complete extract of all open invoices every Sunday night\n- A partner that drops small CSV files into an Amazon S3 bucket several times per hour\n- A website that emits click events continuously to a message bus\n\nWhich mapping of ingestion pattern to source is correct?","options":["Billing extract: streaming ingestion; partner CSVs: reload the entire bucket on every run; click events: weekly scheduled batch","All three sources: nightly batch with full reprocessing, because batch processing keeps track of what has already been processed","Billing extract: batch load; partner CSVs: incremental loading of only the newly arrived files; click events: streaming ingestion","Billing extract: incremental load keyed by invoice ID; partner CSVs: stateful streaming deduplication; click events: hourly batch of the previous hour"],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-engineering/batch-vs-streaming","referenceTitle":"Batch vs. streaming data processing — Databricks Docs"}}
{"id":"RAN-DBDEA-00010","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:b89c614a1bc407bdbc344e744591c60ce37e8849e37d58e1fabe4471bbe62f08","aliases":[],"snapshot":{"statement":"A product manager e-mails the data engineering team a weekly CSV export that lives on her laptop. The team needs its contents in a Unity Catalog-governed Delta table and plans to run this statement once the file is available on a storage path:\n\n```sql\nCOPY INTO main.ops.weekly_orders\nFROM '/Volumes/main/ingest/uploads/'\nFILEFORMAT = CSV;\n```\n\nWhich approach correctly makes the laptop file available to this statement?","options":["Point the FROM clause at the file path on the product manager's laptop, such as 'C:/exports/orders.csv', because COPY INTO can read any path the analyst can open","Upload the file into a Unity Catalog volume (for example, with the Catalog Explorer upload option) so it lands on a UC-governed cloud storage path that COPY INTO can read","Place the file in the workspace DBFS root with dbutils.fs.put, because DBFS paths are automatically governed by Unity Catalog","Attach the CSV to a notebook cell and run spark.read.csv on the uploaded artifact, since notebook attachments behave like Unity Catalog tables"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00011","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:9e48a3731bf82b20fa63e16e5b27de5fe4c338fd0fb6075ec7fff2e77e1efd3e","aliases":[],"snapshot":{"statement":"A data engineer schedules this statement as the single task of a nightly Lakeflow Job. The source folder accumulates Parquet files throughout the day:\n\n```sql\nCOPY INTO main.ops.bookings_raw\nFROM 'abfss://landing@storacct.dfs.core.windows.net/bookings/'\nFILEFORMAT = PARQUET;\n```\n\nWhat happens on a night when no new files have arrived since the last successful run?","options":["The statement completes successfully and loads zero rows, because files that were already loaded by previous runs are skipped","The statement fails with a duplicate-file error, because COPY INTO re-inserts every file in the folder on every run","The statement reloads all files, and the team must schedule a MERGE job afterwards to remove the duplicates","The statement blocks until new files arrive, because COPY INTO behaves like a streaming source waiting for data"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00012","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ab53e9847cc70d3dab1ee628531894d5f8a9ec59f649e32acfb54a557afbd63d","aliases":[],"snapshot":{"statement":"A team must load JSON booking updates from a Unity Catalog volume with COPY INTO. The producer occasionally adds new fields, so the files do not all share the same schema, and the team wants new fields to appear in the target table automatically instead of causing failures. The target table has not been created yet.\n\nWhich approach meets these requirements?","options":["Create the table with an explicit schema listing every column seen so far, so that COPY INTO fails fast whenever a file carries an unexpected field","Create an empty table without declaring columns and set the table property 'delta.columnMapping.mode' = 'name' so unmatched JSON fields map automatically","Read each day's files with spark.read.json, union them with the existing table, and overwrite the table nightly, because COPY INTO cannot handle differing schemas","Create an empty table without declaring columns, then run COPY INTO with 'mergeSchema' = 'true' in FORMAT_OPTIONS and COPY_OPTIONS"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00013","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:af9aca0a22992cc54e02ef9356312d62460b7f90ea48ca3683cee07ba1c73c3c","aliases":[],"snapshot":{"statement":"A newly hired data engineer runs the following statement in a workspace where Unity Catalog manages all data access:\n\n```sql\nCOPY INTO main.ops.shipments\nFROM 'abfss://raw@companylake.dfs.core.windows.net/shipments/'\nFILEFORMAT = PARQUET;\n```\n\nThe target table exists, but the statement fails with a path access error. What must be configured before COPY INTO can read this location?","options":["A DBFS mount point for the storage account, because COPY INTO can only read paths that are mounted under /mnt","A Unity Catalog storage credential and an external location for the path, with READ FILES granted","The storage account key embedded directly in the FROM clause, so COPY INTO authenticates on each run","An ingestion gateway deployed next to the storage account's private endpoint"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00014","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:947a18aade630a32aae717fee78a4d346344edafa33ff7e765f4e7fe09e1780b","aliases":[],"snapshot":{"statement":"A team copies this snippet from an old tutorial to ingest JSON files that keep arriving in a Unity Catalog volume:\n\n```python\ndf = (spark.read\n  .format(\"cloudFiles\")\n  .option(\"cloudFiles.format\", \"json\")\n  .load(\"/Volumes/main/ingest/events/\"))\n```\n\nThe job runs on a schedule, but each run never picks up files that arrived after the very first execution, and no schema inference is configured.\n\nWhich change turns this into a working Auto Loader stream with schema inference and evolution?","options":["Switch the reader to spark.readStream.format(\"cloudFiles\") and keep everything else unchanged, because schema inference needs no additional options","Keep spark.read and add .option(\"cloudFiles.schemaLocation\", ...) to the reader, since the batch reader supports Auto Loader options too","Use spark.readStream.format(\"cloudFiles\") with cloudFiles.format and a durable cloudFiles.schemaLocation for schema inference and evolution","Replace format(\"cloudFiles\") with format(\"streamingFiles\") and set checkpointLocation on the reader, because schemaLocation only applies to Delta targets"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00015","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6014c32e3110facdf435cfcfcf9d684755c8bab30354a6a8209cdbcbdddfcb1d","aliases":[],"snapshot":{"statement":"An Auto Loader stream that feeds a Unity Catalog-governed Delta table failed in the middle of the night after processing part of the incoming file batch. In the morning, the on-call engineer restarts the stream from the same notebook, pointing at the same input directory, checkpoint location, and target table.\n\nWhat does Auto Loader guarantee after the restart?","options":["It resumes where it left off: files already recorded in checkpoint state are not reprocessed, preserving exactly-once semantics","It starts over and re-ingests every file in the input directory, so the on-call engineer must schedule a deduplication job after every recovery","It skips the partially processed files and leaves a gap that must be backfilled manually before the next run","It truncates the target Delta table and reloads all files so the table returns to a consistent state"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00016","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:d9247f48775c5b34f74fa0faec330eb52eff3bdde14faf93ecef7c092940650f","aliases":[],"snapshot":{"statement":"To reduce storage costs, a cloud administrator applies a 30-day lifecycle expiration rule to every prefix under the S3 bucket that hosts an Auto Loader stream's checkpoint directory. The stream has run daily for two months against the same input directory.\n\nWhat is the impact, and what should the team have done instead?","options":["No impact: the checkpoint is rebuilt automatically from the target table's Delta transaction log on the next run","Auto Loader silently switches to file notification mode and re-lists the entire input directory tree on every micro-batch, because its checkpoint state has gone missing","Only files from the last 30 days are reprocessed, because Delta time travel caps recovery at 30 days","The stream state is corrupted and it must restart from scratch, likely reprocessing files; checkpoints belong in durable storage outside lifecycle rules"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/best-practices","referenceTitle":"Auto Loader best practices — Databricks Docs"}}
{"id":"RAN-DBDEA-00017","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ebe55f92d09b04cf389ebac2353646b8f520344dc66ea308b430de709b08e90b","aliases":[],"snapshot":{"statement":"An Auto Loader stream reads files that land in Amazon S3 at a rate that has grown to hundreds of thousands of files per hour spread over a deep directory tree. The stream uses directory listing mode with a continuous trigger, and monthly cloud bills show ever-increasing S3 LIST request charges. The team can obtain the elevated cloud permissions required for notification setup.\n\nWhich change best addresses the cost and scalability problem?","options":["Keep directory listing mode and raise cloudFiles.fetchParallelism so that the directory tree is listed more quickly","Switch to file notification mode: file events go to a queue that the stream consumes, scaling far beyond listing","Require producers to write into a single flat directory so that each run needs only one LIST call","Replace Auto Loader with a scheduled COPY INTO statement, because COPY INTO lists the directory more cheaply than Auto Loader"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/file-notification-mode","referenceTitle":"Auto Loader file notification mode — Databricks Docs"}}
{"id":"RAN-DBDEA-00018","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:c8db53959e2dfefff2f1cd6a90d23247c8e764646c397b12a5f678f2e46d7e33","aliases":[],"snapshot":{"statement":"A team runs Auto Loader against Amazon S3 in classic file notification mode: they set cloudFiles.useNotifications to true and grant Auto Loader the permissions it needs to create cloud resources in the bucket.\n\nWhich resources does Auto Loader create so that newly uploaded files are announced to the stream?","options":["An AWS Lambda function that watches the bucket and writes new object keys to a DynamoDB table that Auto Loader polls","An Amazon Kinesis Data Firehose delivery stream that receives ObjectCreated events and feeds them into an SQS queue","An Amazon SNS topic subscribed to the bucket's object-created notifications, with an Amazon SQS queue that Auto Loader consumes","An Azure Event Grid subscription and an Azure Queue Storage queue, which Databricks uses as its cross-cloud default"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/file-notification-mode","referenceTitle":"Auto Loader file notification mode — Databricks Docs"}}
{"id":"RAN-DBDEA-00019","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:3e05f9c3638f40ba5d7e9174001588378b7a4a5bb86d00fd8d60794e999bb3a8","aliases":[],"snapshot":{"statement":"A bank's security policy forbids creating event notification services or queues in its cloud storage accounts. The data team still wants Auto Loader to ingest new files from a modest source directory that receives a few dozen files per day, with minimal setup.\n\nWhich file discovery mode should the team use, and why?","options":["File notification mode with file events, because managed file events require no cloud permission grants at all","Classic file notification mode, because the team can create the queues themselves instead of letting Auto Loader do it","Plain batch reads with spark.read, because Auto Loader cannot run without notification services","Directory listing mode: the default, needing no cloud permissions beyond read access to the storage path"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/directory-listing-mode","referenceTitle":"Auto Loader directory listing mode — Databricks Docs"}}
{"id":"RAN-DBDEA-00020","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ebf5d93ca177e66d00f38d37d04e93121993cbfe8fbd94f1a13bcf5bb8f67af0","aliases":[],"snapshot":{"statement":"Event files like the one below arrive continuously in a UC volume, and the team starts Auto Loader with schema inference (no explicit schema, cloudFiles.schemaLocation set):\n\n```json\n{\"user_id\": 1042, \"event_ts\": \"2026-05-01T10:22:31Z\", \"amount\": 49.9, \"profile\": {\"tier\": \"gold\", \"region\": \"emea\"}}\n```\n\nWhat schema does Auto Loader infer by default, and how can the team get typed columns such as numbers and timestamps?","options":["All fields — including nested ones — are inferred as string; set cloudFiles.inferColumnTypes to true to infer types like long, double, and timestamp","user_id becomes long, amount becomes double, event_ts becomes timestamp, and profile becomes a struct, because Auto Loader always samples rows and picks the narrowest matching type","The profile object is rejected, because Auto Loader cannot infer nested structures from JSON files","Top-level fields are strings while nested fields keep their JSON types, so only profile.tier and profile.region are correctly typed"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00021","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:a37de25fd62cd8dbd7fbc85f99f45b9e5479b6dc4e12ec490973fde185ab2669","aliases":[],"snapshot":{"statement":"An Auto Loader stream with an inferred schema has run for weeks using only the default schema evolution settings. This morning, the producer adds a new column loyalty_points to the JSON files.\n\nWhat happens on the next micro-batch, and what does Databricks recommend operationally?","options":["The new column is silently ignored forever; to add it, the team must edit the inferred schema files stored under the schema location","The stream keeps running with the old schema and quietly stores all loyalty_points values in the _rescued_data column, with no interruption at all","Auto Loader adds loyalty_points and the stream stops with UnknownFieldException; a restart resumes it with the new schema","The stream drops the offending files into a bad records path and keeps ingesting the remaining files with the old schema"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00022","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:1d1b61f979ca6932f01a1d0b2bb75a736d531ffe47e43e56a5e870aac4c26775","aliases":[],"snapshot":{"statement":"A compliance feed must land in a table whose schema has to stay strictly stable — no automatic column additions. At the same time, the team cannot tolerate ingestion outages when the feed occasionally sends unexpected extra fields, and it wants those values preserved for later inspection.\n\nWhich cloudFiles.schemaEvolutionMode setting satisfies all three requirements?","options":["addNewColumns — new fields are merged into the table schema and the stream resumes automatically","rescue — the schema never evolves, the stream never fails due to schema changes, and unexpected fields are recorded in the rescued data column","failOnNewColumns — the stream stops whenever a new field appears and stays down until someone updates the schema or removes the file","addNewColumnsWithTypeWidening — new fields are merged and compatible type changes are widened automatically"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00023","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c57f443dc095541f6a7bb6f7cee23924495618c041c8628170be1b0107a1a4a1","aliases":[],"snapshot":{"statement":"An Auto Loader stream runs with an inferred schema. In today's batch, some records use a column name with different casing (OrderID instead of order_id), and one field's value does not match the type Auto Loader inferred for it.\n\nWhere does this non-conforming data end up?","options":["The affected records are dropped silently; only a row-count discrepancy in the stream metrics hints that something was lost","The stream fails immediately, because case mismatches and type mismatches are fatal errors during schema inference","The mismatched values are redirected to the bad records path that Auto Loader configures by default for every stream","They land in the _rescued_data column Auto Loader adds automatically — non-conforming fields plus the source file path"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00024","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a92322014af65449b4104877f460cce88e62da9aaf06bf51c193b3aaf9aae3cc","aliases":[],"snapshot":{"statement":"Auto Loader is ingesting the JSON events below with schema inference, so every field — nested ones included — is currently a string:\n\n```json\n{\"event_ts\": \"2026-05-01T10:22:31Z\", \"amount\": \"49.90\", \"user_info\": {\"id\": \"1042\", \"dob\": \"1990-04-18\"}}\n```\n\nThe team wants event_ts read as TIMESTAMP, amount as DOUBLE, and user_info.dob as DATE, while all other fields keep being inferred automatically.\n\nWhich Auto Loader option does this?","options":["cloudFiles.schemaHints, e.g. 'event_ts TIMESTAMP, amount DOUBLE, user_info.dob DATE' — pins known types, infers the rest","cloudFiles.schemaEvolutionMode set to addNewColumnsWithTypeWidening, which converts inferred strings into native types as data arrives","cloudFiles.inferColumnTypes together with a full .schema() declaration listing every column, including all nested fields","cloudFiles.rescuedDataColumn, which casts values rescued from type mismatches back into typed columns"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00025","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:4ba7b1f24884c331084238b78de942a66aa89fe466a01230d9f394014de1156d","aliases":[],"snapshot":{"statement":"A retailer wants to ingest files from cloud object storage into a Unity Catalog-governed table once per night. The pipeline should start on a schedule, process whatever files have arrived, and then stop, so the team pays only for a short run instead of compute that stays on around the clock.\n\nWhich design matches these requirements?","options":["Start a stream with .trigger(processingTime='1 hour') and leave it running continuously, so each micro-batch picks up new files as they arrive","Use .trigger(availableNow=True): process all available data, then stop — scheduled via a time-based trigger in a Lakeflow Job","Run Trigger.Once inside a while loop in a single notebook so the stream never gets a chance to idle","Skip Auto Loader entirely and use a one-time spark.read.parquet, because Auto Loader cannot be used for scheduled batch-style loads"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/best-practices","referenceTitle":"Auto Loader best practices — Databricks Docs"}}
{"id":"RAN-DBDEA-00026","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a530e635f75db6f689c2bcbd06a625cd75f4d8e4c893a5c580660c6df7b66744","aliases":[],"snapshot":{"statement":"A logistics company must ingest shipment telemetry from a supplier's bucket: roughly 5,000 new Parquet files per day arriving around the clock, volume growing about 20% per year, and the producer occasionally adds columns. The team wants low operational overhead with schema handled automatically. The SQL-first team lead proposes a scheduled COPY INTO statement that runs every 15 minutes.\n\nWhat should the data engineering team recommend?","options":["Accept COPY INTO — it scales to any file count and absorbs column changes as long as it runs frequently enough","Replace it with a nightly spark.read.parquet overwrite of the entire bucket, which avoids streaming checkpoints entirely","Use Auto Loader (e.g. a streaming table): incremental file tracking, very large scale, and schema inference with evolution","Query the supplier's operational database over JDBC instead, because file-based ingestion cannot keep up with around-the-clock arrivals"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00027","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:190ef668e5ac1c917a2b57f9f1d00b9ba52d478a049d5a59642d2a8927c0793c","aliases":[],"snapshot":{"statement":"Sales operations needs Salesforce data — accounts, opportunities, and their ongoing changes — replicated into Unity Catalog-governed Delta tables. The requirements are minimal operational overhead, no ingestion code for the team to maintain, change capture from the SaaS source, and full governance on the target tables.\n\nWhich approach should the data engineering team prioritize?","options":["A nightly full CSV export from Salesforce into S3, loaded by a scheduled COPY INTO statement","A partner ELT tool deployed on the team's own compute that writes Parquet files to an external location","A custom Structured Streaming job against the Salesforce REST API, owned and maintained by the platform team, with the team patching and scaling its compute","The Lakeflow Connect managed Salesforce connector: serverless, Databricks-operated, landing in UC-governed streaming tables"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ingestion/lakeflow-connect/","referenceTitle":"Lakeflow Connect connector concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00028","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:bf3f278c77969730626d1e026a6acdedb4c63ac97473274505cc029380f677d8","aliases":[],"snapshot":{"statement":"The team configures a Lakeflow Connect managed streaming connector for a Confluent Kafka topic that carries semi-structured JSON order events. Everything must live in Unity Catalog, and the team must not manage any compute for the ingestion.\n\nWhich set of components does Lakeflow Connect stand up, and how does the data flow?","options":["A Unity Catalog connection for credentials, a serverless pipeline, and UC-governed streaming tables as destination","A job cluster that runs an Auto Loader notebook and writes Parquet files to an external location registered as a foreign catalog","An ingestion gateway virtual machine in the team's VPC plus staging storage that the team must size, patch, and monitor","A Lakeflow Job with a dbt task that materializes the topic into hourly materialized views"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/lakeflow-connect/","referenceTitle":"Lakeflow Connect connector concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00029","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:8b953004ed5c8e251852be7a260d24026d9db3072f5c6dc6688e3138646c858c","aliases":[],"snapshot":{"statement":"A notebook task extracts a 400-million-row table from a PostgreSQL data warehouse over JDBC. The current read runs on a single connection and takes hours:\n\n```python\norders = (spark.read.format(\"jdbc\")\n  .option(\"url\", jdbc_url)\n  .option(\"dbtable\", \"sales.orders\")\n  .option(\"user\", user)\n  .option(\"password\", pwd)\n  .load())\n```\n\nHow should the team parallelize the read, and what is the effect?","options":["Set numPartitions=8 alone; Spark opens eight simultaneous queries against sales.orders without needing any column or bound options","Add partitionColumn (indexed, uniform), lowerBound, upperBound, and numPartitions; Spark runs concurrent range queries","Set fetchSize=100000 so each round trip pulls more rows through the existing single connection, which parallelizes the fetch","Set partitionColumn to the primary key and numPartitions=500 to maximize the number of simultaneous queries against the warehouse"],"answer":"B","reference":"https://docs.databricks.com/aws/en/archive/connectors/jdbc","referenceTitle":"Query databases using JDBC — Databricks Docs"}}
{"id":"RAN-DBDEA-00030","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:be68e040286f588541c607763549ba1dc913dbfe860d4949c3c36b9522a76b2b","aliases":[],"snapshot":{"statement":"A vendor exposes order data only through a paged REST API that returns JSON documents; there are no file drops and no direct database access. The team must land this data daily into a Unity Catalog-governed table. A junior engineer proposes:\n\n```sql\nCOPY INTO main.ops.orders\nFROM 'https://api.vendor.com/v2/orders'\nFILEFORMAT = JSON;\n```\n\nWhat is the correct approach?","options":["Run the proposed statement with FORMAT_OPTIONS ('header' = 'true') added, so that the HTTP endpoint becomes readable as JSON","Ask the vendor to publish the data to Kafka, because REST API sources cannot be landed in Unity Catalog tables","Register the API URL as a Unity Catalog external location so that COPY INTO treats it as cloud object storage","Write a notebook that pages through the API and appends to a UC-governed table, scheduled as a Lakeflow Job task"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/","referenceTitle":"Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00031","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d064eeaa2dfbaf88bb2b78c476f3a29b5c846148a8d084d930ead18bacceb2b4","aliases":[],"snapshot":{"statement":"A junior engineer tries to start an incremental read of a Kafka topic called orders on a Confluent cluster:\n\n```python\ndf = (spark.read\n  .format(\"confluent\")\n  .option(\"bootstrap\", \"broker1:9092\")\n  .option(\"topic\", \"orders\")\n  .load())\n```\n\nThe call fails immediately. Which correction correctly starts the streaming read so that the JSON message bodies can be parsed into a Delta table?","options":["Use spark.readStream.format(\"kafka\") with kafka.bootstrap.servers and subscribe set to \"orders\"; parse the binary columns downstream","Keep spark.read and just change the format to \"kafka\"; batch reads automatically become incremental once the result is written to a Delta table","Use spark.readStream.format(\"cloudFiles\") with cloudFiles.format set to \"kafka\" so that Auto Loader discovers new messages as they arrive","Use spark.readStream.format(\"jdbc\") with the broker URL, because Kafka exposes a JDBC interface for external consumers"],"answer":"A","reference":"https://docs.databricks.com/aws/en/connect/streaming/kafka/","referenceTitle":"Connect to Apache Kafka — Databricks Docs"}}
{"id":"RAN-DBDEA-00032","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:93d3686385f2f92863760a9c5a97bd12b068aca2d6140d91495f5b7aae8b3211","aliases":[],"snapshot":{"statement":"A company is building its lakehouse on the medallion architecture. Several sources feed the first layer: Parquet files in Amazon S3, events from Kafka, and Salesforce data through Lakeflow Connect. Before any transformation is applied, the engineers must agree on what this first layer should contain.\n\nWhich statement best describes what belongs in that first layer?","options":["Joins and aggregations across all the sources, so that downstream layers only need lightweight filtering to serve consumers","Cleaned, deduplicated, and validated records, so later layers can assume the data is already correct","The raw source data landed as-is and appended incrementally, preserving the original state as the source of truth","Aggregated KPIs for executives, while the raw data stays in an archive bucket outside Unity Catalog"],"answer":"C","reference":"https://docs.databricks.com/aws/en/lakehouse/medallion","referenceTitle":"Medallion architecture — Databricks Docs"}}
{"id":"RAN-DBDEA-00033","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6033e52286c64ab646db51a7411b4b21246610411dc2e220efd9fd0ad915e16a","aliases":[],"snapshot":{"statement":"A data engineer is building the silver layer of a medallion pipeline. The bronze table catalog.sales.bronze_customers contains a nullable STRING column named region. For catalog.sales.silver_customers, every customer row must be kept, but region must never be NULL in silver — rows without a region must show the value \"Unknown\".\n\n```python\nbronze = spark.read.table(\"catalog.sales.bronze_customers\")\nsilver = <cleaning step>\nsilver.write.saveAsTable(\"catalog.sales.silver_customers\")\n```\n\nWhich cleaning step meets this requirement?","options":["silver = bronze.fillna('Unknown', subset=['region'])","silver = bronze.fillna(0, subset=['region'])","silver = bronze.dropna(subset=['region'])","silver = bronze.withColumn('region', lit('Unknown'))"],"answer":"A","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00034","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:60cee13e00127038d49dfe2e3356b547f553af8fa5f5e4a89caca43d566ff1da","aliases":[],"snapshot":{"statement":"The bronze table catalog.crm.bronze_contacts stores two email columns. primary_email is frequently NULL, and some of those rows have a usable secondary_email. The silver view must expose a single column contact_email that returns primary_email when it is not NULL, otherwise secondary_email, and otherwise the literal 'no-email@company.com'.\n\n```sql\nSELECT customer_id,\n       /* expression here */ AS contact_email\nFROM catalog.crm.bronze_contacts\n```\n\nWhich expression implements this fallback logic?","options":["NULLIF(primary_email, secondary_email)","COALESCE('no-email@company.com', primary_email, secondary_email)","COALESCE(primary_email, secondary_email, 'no-email@company.com')","NVL2(primary_email, secondary_email, 'no-email@company.com')"],"answer":"C","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00035","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:72e230e9d27a757eedc8d5be01add13a180fbb025603f94b60c6a68b7886bcc1","aliases":[],"snapshot":{"statement":"An events bronze table landed event_time as STRING values like '2026-03-14 09:30:00' and amount as STRING values like '125.50'. The silver table must standardize the types: event_time as TIMESTAMP and amount as DOUBLE, keeping the same column names and all other columns unchanged.\n\n```python\nbronze = spark.read.table(\"catalog.ops.bronze_events\")\nsilver = <type standardization step>\nsilver.write.saveAsTable(\"catalog.ops.silver_events\")\n```\n\nWhich step standardizes the data types correctly?","options":["silver = bronze.withColumnRenamed('event_time', 'event_time_ts').withColumnRenamed('amount', 'amount_double').withColumnRenamed('cust', 'customer')","bronze.withColumn('event_time', col('event_time').cast('timestamp')).withColumn('amount', col('amount').cast('double'))","silver = bronze.select(col('event_time').cast('date'), col('amount').cast('double'))","silver = bronze.withColumn('event_time', to_date('event_time'))"],"answer":"B","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00036","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:e9aae7f92dd7a0a98b97f804a3d225185f00be98e924166cc518b9df21cbc5a2","aliases":[],"snapshot":{"statement":"During ingestion design, a junior engineer proposes dropping rows with NULLs and casting every column to its final type before writing to the bronze table, arguing that the silver pipelines would then \"just work\" with no cleaning code at all. A senior engineer objects, citing the responsibilities of the medallion architecture layers.\n\nWhere does this cleaning work belong, and why?","options":["In the bronze layer, because bronze must store only validated, strongly typed data before any table is published","In the gold layer, because only gold tables are schema-enforced in Unity Catalog","In the source system, because the medallion architecture requires clean data before any ingestion","In the silver layer — bronze should keep raw data as landed so it can be replayed and audited"],"answer":"D","reference":"https://docs.databricks.com/aws/en/lakehouse/medallion","referenceTitle":"Medallion lakehouse architecture — Databricks Docs"}}
{"id":"RAN-DBDEA-00037","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f38189552c726cdd0db8b20b4731677c6f0c4756c96a799416a8ca093e3ac230","aliases":[],"snapshot":{"statement":"A silver table of customer contacts is usable as long as a customer can be reached by at least one channel. Rows where BOTH email and phone are NULL must be removed. Rows where only one of the two is NULL must be kept, with the NULL value preserved in place.\n\n```python\nbronze = spark.read.table(\"catalog.crm.bronze_contacts\")\nsilver = <drop rule>\n```\n\nWhich dropna call implements this rule?","options":["silver = bronze.dropna(how='any', subset=['email', 'phone'])","silver = bronze.dropna(how='all', subset=['email', 'phone'])","silver = bronze.dropna(subset=['email'])","silver = bronze.dropna(thresh=2)"],"answer":"B","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00038","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:620ebd8d8c25d987bccf7e6016e839676c6c46760c7b60d4fee5032a85a270ce","aliases":[],"snapshot":{"statement":"While validating a silver notebook, an engineer joins two small bronze tables to check key overlap before promoting the pipeline:\n\n```python\npayments = spark.read.table(\"catalog.fin.bronze_payments\")   # 3 rows: txn 101, 202, 303\nrefunds   = spark.read.table(\"catalog.fin.bronze_refunds\")   # 4 rows: txn 202, 303, 404, 505\n\njoined = payments.join(refunds, \"txn_id\", \"inner\")\nprint(joined.count())\n```\n\nWhat number is printed, and why?","options":["2 — only the txn_id values present in both tables survive the inner join","3 — all payments are preserved because the left table drives an inner join","4 — all refunds are preserved because the right table drives an inner join","7 — an inner join stacks both inputs and removes nothing"],"answer":"A","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00039","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:5b8f6346648cba6481532491184b82f42c6f66816a3cc252fa3a062393940d1e","aliases":[],"snapshot":{"statement":"A daily silver job joins about 1 million sales rows to dim_prices, which holds one row per (product_id, store_id) because prices vary by store. After the join runs, the silver table contains about 3 million rows and revenue totals are inflated. The suspicious line:\n\n```python\nsilver = (sales\n    .join(dim_prices, \"product_id\")\n    .withColumn(\"revenue\", col(\"qty\") * col(\"price\")))\n```\n\nWhat is the most likely cause, and the correct fix?","options":["The sales fact contains duplicate transactions — call dropDuplicates() with no subset before the join","Spark picked the wrong physical join strategy — force a broadcast join with .hint('broadcast') so each sale matches exactly one price","The join key is not unique in dim_prices, so sales fan out one row per store — join on both product_id and store_id","dim_prices contains NULL prices that replicate rows during the join — filter out NULL prices before joining"],"answer":"C","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00040","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:cc0a37a70b4d880823cd5f01d614f1cd02feab5953f2fcbaf403740d10a534df","aliases":[],"snapshot":{"statement":"A nightly job joins a 900 GB transactions DataFrame to a 250 MB currency_rates dimension on currency_code and then aggregates by market. The Spark UI shows a SortMergeJoin whose shuffle of the transactions data dominates the runtime. The rates table fits comfortably in every executor's memory.\n\n```python\njoined = txns.join(rates, \"currency_code\")\n```\n\nWhich change most directly avoids shuffling the large DataFrame for this join?","options":["Increase spark.sql.shuffle.partitions so each task shuffles a smaller slice of the transactions data","Call txns.cache() before the join so Spark reuses a shuffle-free copy of the transactions data","Run OPTIMIZE on the currency_rates table to compact its files before the join","Force a broadcast hash join by wrapping the small side: txns.join(broadcast(rates), 'currency_code')"],"answer":"D","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00041","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:400b196ca3220f949f0c068d992a080ee7e6ad26a54b916ff991fc1b50c696ee","aliases":[],"snapshot":{"statement":"A marketing engineer needs every combination of customer segment (40 rows) and promo code (5,000 rows) for a campaign evaluation job, and writes:\n\n```python\nresult = segments.crossJoin(promos)\n```\n\nThe job will score every resulting pair once per night. What does this produce, and what should the engineer keep in mind?","options":["A 200,000-row Cartesian product (40 x 5,000) — fine here, but cross joins grow as rows-times-rows","5,040 rows — crossJoin stacks both DataFrames the same way a union does","An error, because cross joins require setting spark.sql.crossJoin.enabled to true first","Only the rows where a segment value also appears in the promos DataFrame — crossJoin applies an implicit equality filter"],"answer":"A","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00042","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:56a7d9322bd34bd835b64041b688711893f3acd24fac2009cf6cfa5345214796","aliases":[],"snapshot":{"statement":"Two regional extracts can overlap: the same order row may appear in both files. An engineer combines them expecting SQL UNION semantics, where duplicate rows collapse into one:\n\n```python\ncombined = extract_eu.union(extract_us)\n```\n\nThe result still contains duplicated rows. Why, and how should the code be fixed?","options":["DataFrame.union() deduplicates only when both sides have identical column names — rename the columns first","union() is UNION ALL by design — rows stack positionally; append .distinct() to deduplicate","unionAll() is the deduplicating variant — switch to extract_eu.unionAll(extract_us)","union() runs asynchronously, so duplicates appear until the result is materialized — call .cache() and count() first"],"answer":"B","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00043","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:91a5903bd6370fd72f553fd2eed2b5261db2acb39b7f66450ad909a1227d5421","aliases":[],"snapshot":{"statement":"Building catalog.sales.silver_customers from bronze, an engineer must: (1) split the STRING column full_name on the single space character into two new STRING columns, first_name and last_name; (2) drop full_name and the PII column ssn; (3) rename cust_id to customer_id.\n\nA teammate proposes:\n\n```python\nsilver = (bronze\n    .withColumn(\"first_name\", split(col(\"full_name\"), \" \"))\n    .withColumn(\"last_name\", split(col(\"full_name\"), \" \"))\n    .drop(\"full_name\", \"ssn\")\n    .withColumnRenamed(\"cust_id\", \"customer_id\"))\n```\n\nWhat is wrong with this proposal?","options":["Nothing is wrong — the chain meets all three requirements as written","drop() cannot remove two columns in a single call — full_name and ssn must be dropped in separate drop() calls","split() returns an ARRAY, so the new columns would hold arrays — use .getItem(0) and .getItem(1)","withColumnRenamed('cust_id', 'customer_id') renames in the wrong direction — the arguments are (new, existing)"],"answer":"C","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00044","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:7f5d893c5a100fc03c3eafb5d2b8ae58adcfbbab3c03255f8713d4670c799967","aliases":[],"snapshot":{"statement":"The bronze table catalog.web.bronze_search stores an ARRAY<STRING> column tags on each search event. For tag-level analytics, the silver table needs one row per (event_id, tag) pair — an event carrying three tags becomes three rows.\n\n```sql\nSELECT event_id, /* expression */ AS tag\nFROM catalog.web.bronze_search\n```\n\nWhich expression produces one row per array element?","options":["explode(tags)","flatten(tags)","collect_list(tags)","array_distinct(tags)"],"answer":"A","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00045","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:4ec09c5db3e40ebd5aa3eb109c3fc9bd99e44f58fd7a0bd7fba043489baa59fb","aliases":[],"snapshot":{"statement":"A silver pipeline must exclude cancelled orders but keep every other row, including orders whose status is still NULL because they are pending manual review. After running:\n\n```python\nsilver = bronze.filter(col(\"status\") != \"CANCELLED\")\n```\n\nQA reports that 612 rows are missing from silver — precisely the rows with NULL status. Why did this happen, and what is the correct fix?","options":["filter() silently ignores rows with NULL in the filtered column — .where() handles NULL predicates correctly, so switch to .where()","The Spark reader dropped the NULL statuses at load time — reload the table with .option('mode', 'PERMISSIVE')","Fill the NULLs before filtering: bronze.fillna('UNKNOWN', subset=['status']) makes every row comparable, so the filter keeps them","The predicate is NULL for NULL statuses and filter() keeps only true — add an isNull() branch or a null-safe operator"],"answer":"D","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00046","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:35acd2b5fb033575134e7fe0b184fdff81f295c3065126374254f6e0a8d4cd61","aliases":[],"snapshot":{"statement":"bronze_orders accumulates re-delivered files: the same order_id appears multiple times with different ingested_at values, and silver must keep exactly the most recent row per order. A teammate proposes:\n\n```python\nfrom pyspark.sql.functions import col\n\nlatest = (bronze\n    .orderBy(col(\"ingested_at\").desc())\n    .dropDuplicates([\"order_id\"]))\n```\n\nWhat is the correct assessment of this proposal?","options":["The proposal is correct — orderBy(col('ingested_at').desc()) guarantees that dropDuplicates(['order_id']) keeps the newest row per order, because the sort carries over","dropDuplicates(['order_id']) keeps an arbitrary row, ignoring the sort — use row_number() over a window ordered by ingested_at DESC, keeping rn = 1","dropDuplicates(['order_id']) is missing columns — listing every column of the table would make it keep only the latest row","Replace dropDuplicates(['order_id']) with distinct(), which keeps the last row encountered after an orderBy"],"answer":"B","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00047","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:63a524a47ef104cdcd8fbc16edf79471eb5c640d9320bd2c8ca3abab9e62c011","aliases":[],"snapshot":{"statement":"While validating a silver table, an engineer wants a one-call profile of the numeric column amount: count, mean, standard deviation, min, max, and the approximate 25th, 50th, and 75th percentiles — without writing any custom aggregation code.\n\nWhich DataFrame method returns all of those statistics?","options":["df.describe('amount')","df.explain()","df.printSchema()","df.summary()"],"answer":"D","reference":"https://docs.databricks.com/aws/en/pyspark/reference/functions","referenceTitle":"PySpark function reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00048","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:fd40af1ba7859c0eb477ac7e85fe031810da158655f2e574f38ac7cc51fb5c54","aliases":[],"snapshot":{"statement":"A daily gold KPI reports distinct visitors per date over roughly 800 million events. The exact query below runs for about 25 minutes because counting distinct values forces a massive shuffle to bring all values of each key together:\n\n```sql\nSELECT event_date, count(DISTINCT visitor_id) AS visitors\nFROM catalog.web.silver_events\nGROUP BY event_date;\n```\n\nThe dashboard tolerates an error of a few percent in exchange for a much faster query. Which function should replace count(DISTINCT ...)?","options":["count(visitor_id) — counting non-null rows is statistically equivalent to counting unique visitors","size(collect_set(visitor_id)) — collecting the exact set of visitors per date","approx_count_distinct(visitor_id) — a sketch-based approximate distinct count with far less shuffling","percentile_approx(visitor_id, 0.5) — the approximate median visitor id"],"answer":"C","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00049","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:fe57bc0576415403cb85383a5c43b9c3f28dd689cb20d47b487df68d0116a6c6","aliases":[],"snapshot":{"statement":"Every night, catalog.ops.order_updates arrives with changed rows for existing order_id keys plus brand-new orders. catalog.ops.silver_orders must reflect the updates in place and contain the new orders — with no duplicate order_id values. All tables are Delta tables governed by Unity Catalog.\n\nWhich SQL statement is designed for this requirement?","options":["MERGE INTO catalog.ops.silver_orders AS t USING catalog.ops.order_updates AS s ON t.order_id = s.order_id WHEN MATCHED THEN UPDATE SET t.status = s.status, t.amount = s.amount WHEN NOT MATCHED THEN INSERT *","INSERT INTO catalog.ops.silver_orders SELECT * FROM catalog.ops.order_updates","DELETE FROM catalog.ops.silver_orders WHERE order_id IN (SELECT order_id FROM catalog.ops.order_updates); then INSERT INTO catalog.ops.silver_orders SELECT * FROM catalog.ops.order_updates","CREATE OR REPLACE TABLE catalog.ops.silver_orders AS SELECT * FROM catalog.ops.silver_orders UNION ALL SELECT * FROM catalog.ops.order_updates"],"answer":"A","reference":"https://docs.databricks.com/aws/en/delta/merge","referenceTitle":"Merge data (Delta Lake upsert) — Databricks Docs"}}
{"id":"RAN-DBDEA-00050","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ffe5561543322914d3b1e27f77fa3900a20e5fde729792bbe1b720da7dd96f2b","aliases":[],"snapshot":{"statement":"On a cluster with 4 workers, a silver aggregation over a 3 GB DataFrame shows 200 shuffle tasks in the Spark UI; most tasks process only a few megabytes, and scheduling overhead dominates the stage time. The team wants fewer, larger tasks. spark.sql.shuffle.partitions is at its default value.\n\nWhich statement is accurate?","options":["spark.default.parallelism sets the shuffle partition count for SQL aggregations — lower it to 32 on this cluster","spark.sql.shuffle.partitions (default 200) sets the shuffle partition count — lower it or rely on AQE coalescing","spark.sql.autoBroadcastJoinThreshold caps the byte size of each shuffle partition — raise it so partitions grow larger","spark.executor.memory limits how many tasks an executor can schedule — increase it to reduce the task count"],"answer":"B","reference":"https://docs.databricks.com/aws/en/optimizations/aqe","referenceTitle":"Adaptive query execution — Databricks Docs"}}
{"id":"RAN-DBDEA-00051","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0ce244210938f710107bdaaef25fdafa5a6a943813a0b9c6c132026ea7bcef65","aliases":[],"snapshot":{"statement":"A data engineer compares a query's static plan with its executed plan: a SortMergeJoin became a BroadcastHashJoin, and the number of shuffle partitions in a later stage dropped from 200 to 12 after the first stage completed. The team debates what caused these changes.\n\nWhich statement about Adaptive Query Execution (AQE) is correct?","options":["AQE caches intermediate results in Delta Lake and replays them whenever the same query re-runs","AQE only optimizes queries whose source tables are cached in memory — uncached tables always use the static plan, with no runtime re-optimization","AQE re-optimizes at runtime using statistics from completed stages, coalescing partitions and converting sort-merge to broadcast joins","AQE is disabled by default and must be enabled per query with the .hint('aqe') table hint"],"answer":"C","reference":"https://docs.databricks.com/aws/en/optimizations/aqe","referenceTitle":"Adaptive query execution — Databricks Docs"}}
{"id":"RAN-DBDEA-00052","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9a0a84fb08b60843ebced674fb51c897b9978048413ba5a8787ddd3475817d5d","aliases":[],"snapshot":{"statement":"A shuffle-heavy silver aggregation repeatedly fails with a java.lang.OutOfMemoryError raised inside an executor while tasks build their aggregation buffers; the driver stays healthy. The compute runs a worker configuration with a small heap per executor.\n\nWhich configuration change addresses the root cause most directly?","options":["Increase executor memory (larger worker type or higher spark.executor.memory)","Increase spark.driver.memory so the driver can buffer bigger aggregation results","Increase spark.sql.autoBroadcastJoinThreshold so the large table is broadcast to every executor","Decrease spark.executor.memory so Spark automatically splits oversized tasks"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Compute configuration reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00053","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:65e5ee5465a9766342de809ab94579ab73eae540df65f60e8b65c9d3a062aae8","aliases":[],"snapshot":{"statement":"A BI dashboard queries a gold aggregation — complex joins plus a GROUP BY over billions of rows — on every page load, and each query takes about 40 seconds. The data only needs to be current as of the last refresh, and the team refreshes analytics hourly. They want a Unity Catalog object whose results are precomputed and stored so that dashboard reads are fast.\n\nWhich object should the dashboard query?","options":["A standard view, because views persist query results between accesses","A materialized view, refreshed hourly — precomputed and stored for fast reads","A streaming table, because streaming tables precompute dashboard aggregations from batch tables","A temporary view, because temporary views are visible to every BI session in the workspace"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ldp/concepts/materialized-views","referenceTitle":"Materialized views (Lakeflow pipelines) — Databricks Docs"}}
{"id":"RAN-DBDEA-00054","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:c830430b7c2f1a562949dc13b57c32afea8712278797b034dd0ef69003b77714","aliases":[],"snapshot":{"statement":"The analytics team wants catalog.analytics.daily_revenue as a precomputed daily revenue rollup over catalog.sales.gold_orders. The underlying aggregation is expensive, so recomputing it on every dashboard query is unacceptable; BI tools must read the object like a normal table, and the results must be refreshed automatically every night at 02:00 UTC with no manual steps.\n\nWhich set of statements achieves this?","options":["CREATE VIEW catalog.analytics.daily_revenue AS SELECT order_date, sum(amount) AS revenue FROM catalog.sales.gold_orders GROUP BY order_date; ALTER VIEW catalog.analytics.daily_revenue ADD SCHEDULE CRON '0 0 2 * * ?' AT TIME ZONE 'UTC';","CREATE OR REFRESH STREAMING TABLE catalog.analytics.daily_revenue SCHEDULE CRON '0 0 2 * * ?' AT TIME ZONE 'UTC' AS SELECT order_date, sum(amount) AS revenue FROM catalog.sales.gold_orders GROUP BY order_date;","CREATE TABLE catalog.analytics.daily_revenue TBLPROPERTIES ('refresh.cron' = '0 0 2 * * ?') AS SELECT order_date, sum(amount) AS revenue FROM catalog.sales.gold_orders GROUP BY order_date;","CREATE MATERIALIZED VIEW catalog.analytics.daily_revenue SCHEDULE CRON '0 0 2 * * ?' AT TIME ZONE 'UTC' AS SELECT order_date, sum(amount) AS revenue FROM catalog.sales.gold_orders GROUP BY order_date;"],"answer":"D","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-syntax-ddl-create-materialized-view","referenceTitle":"CREATE MATERIALIZED VIEW — Databricks Docs"}}
{"id":"RAN-DBDEA-00055","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:03a3988b94b13b5a041d150f991bbb80714de491aa1c9f0f045a23082da79b93","aliases":[],"snapshot":{"statement":"JSON event files land continuously in cloud storage. The gold table catalog.analytics.gold_site_events must be built incrementally inside a Lakeflow Spark Declarative Pipeline: every refresh must process only the newly arrived files, exactly once, with the pipeline tracking how far it has read.\n\nWhich declaration meets these requirements?","options":["CREATE MATERIALIZED VIEW catalog.analytics.gold_site_events AS SELECT * FROM STREAM(read_files('s3://company-raw/site_events/', format => 'json'));","CREATE VIEW catalog.analytics.gold_site_events AS SELECT * FROM STREAM(read_files('s3://company-raw/site_events/', format => 'json'));","CREATE TABLE catalog.analytics.gold_site_events AS SELECT * FROM read_files('s3://company-raw/site_events/', format => 'json'); — then a scheduled job runs INSERT INTO ... SELECT * FROM read_files(...) to append new files","CREATE OR REFRESH STREAMING TABLE catalog.analytics.gold_site_events AS SELECT * FROM STREAM(read_files('s3://company-raw/site_events/', format => 'json'));"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ldp/concepts/streaming-tables","referenceTitle":"Streaming tables (Lakeflow pipelines) — Databricks Docs"}}
{"id":"RAN-DBDEA-00056","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:75b35a2575d47acbc9edfe5675004e5b00fa096a8adb599c9c6ba9f43578149c","aliases":[],"snapshot":{"statement":"A team hardening silver tables in Unity Catalog debates how far the database itself will enforce data quality:\n\n- \"If we add a PRIMARY KEY, duplicate keys can never be inserted.\"\n- \"CHECK constraints only warn the writer; the row is still committed.\"\n- \"NOT NULL can only be set at CREATE TABLE time.\"\n\nWhich statement about table constraints in Databricks is correct?","options":["CHECK constraints are advisory — they are recorded in the table schema but never block or roll back a write","PRIMARY KEY constraints reject INSERT statements that would create duplicate key values","NOT NULL and CHECK are enforced on writes; PRIMARY KEY and FOREIGN KEY are informational only","Constraints apply only to external tables; managed tables skip validation entirely"],"answer":"C","reference":"https://docs.databricks.com/aws/en/tables/constraints","referenceTitle":"Table constraints — Databricks Docs"}}
{"id":"RAN-DBDEA-00057","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9d544759977df437c22248f7aacba2f04064a6622b53e638f6023f4b70f197d9","aliases":[],"snapshot":{"statement":"The analytics team defines a gold materialized view in SQL over silver tables. Its data quality rule: margin_pct must be between 0 and 100. Rows that violate the rule must be excluded from the materialized view, and the number of violations must still be recorded as expectation metrics in the event log. The refresh itself must not fail because of bad rows.\n\nWhich clause implements this rule?","options":["CONSTRAINT valid_margin EXPECT (margin_pct BETWEEN 0 AND 100) ON VIOLATION FAIL UPDATE","CONSTRAINT valid_margin EXPECT (margin_pct BETWEEN 0 AND 100) ON VIOLATION DROP ROW","CONSTRAINT valid_margin EXPECT (margin_pct BETWEEN 0 AND 100)","CONSTRAINT valid_margin CHECK (margin_pct BETWEEN 0 AND 100)"],"answer":"B","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-syntax-ddl-create-materialized-view","referenceTitle":"CREATE MATERIALIZED VIEW — Databricks Docs"}}
{"id":"RAN-DBDEA-00058","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:09ca4ea654edc6c17e316610eb9f47aeb980729ea4ab51cfbcd88bb444c42c7e","aliases":[],"snapshot":{"statement":"A Lakeflow Job contains four tasks: `ingest` (notebook task), `validate` (SQL task), `enrich` (pipeline task), and `publish` (dashboard task). `validate` and `enrich` each depend only on `ingest`. `publish` depends on both `validate` and `enrich`, and every dependency uses the default run-if condition. In the latest run, `ingest` succeeded, but `validate` failed.\n\nWhat is the expected behavior?","options":["The run is canceled immediately, so enrich and publish never start","Enrich runs, and publish also runs because one of its two dependencies succeeded","Enrich runs — it depends only on ingest; publish is skipped since a dependency failed","Both enrich and publish are skipped, because Lakeflow Jobs stops every task in the run after any single task fails"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/control-flow","referenceTitle":"Control the flow of tasks within Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00059","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b7fa94ae8adad3d4b3f22e986fbf23d3e29e0b1dfa20fa58517dceb2c47cad3c","aliases":[],"snapshot":{"statement":"A nightly Lakeflow Job loads data with several tasks. No matter whether those loading tasks succeed or fail, the team must always archive the temporary landing directory at the end of the run. They plan to add a final cleanup task that depends on all loading tasks.\n\nWhich run-if dependency should the cleanup task use?","options":["All succeeded, so the cleanup runs after every dependency completes without error","At least one succeeded, so the cleanup runs whenever any loading task finishes successfully, even if others are still running","None failed, which is the only condition that guarantees the cleanup task always runs","All done — the cleanup runs once its dependencies finish, regardless of their outcomes"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/control-flow","referenceTitle":"Control the flow of tasks within Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00060","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c1e1886725aa0b210f0046a701d7d31450b2df948b060e3fb34380236f4534fe","aliases":[],"snapshot":{"statement":"A notebook task in a Lakeflow Job calls a partner REST API that intermittently returns HTTP 503 errors. The team wants the task to be re-run automatically up to three times before it is marked as failed. The task also has a timeout configured.\n\nHow should the team meet this requirement?","options":["In the task's advanced settings, click Add next to Retries and set the retry count and interval","Set the job's Maximum concurrent runs to 3, so failed attempts are replaced by fresh runs","Wrap the notebook in a For each task with three iterations that re-invokes it on failure","Configure a failure notification destination that triggers Run now whenever a failure email arrives"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/configure-task","referenceTitle":"Configure and edit tasks in Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00061","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ac2d49dd36703a5960ccfa9449e7be90ec750a89bf49610eed88d77b9e764e3d","aliases":[],"snapshot":{"statement":"The first task in a Lakeflow Job, named qc, ends with `dbutils.jobs.taskValues.set(key = \"bad_records\", value = bad_count)`. If bad_records is greater than zero, the quarantine notebook must run; otherwise the publish notebook must run.\n\nHow should the data engineer implement this branching?","options":["Add a For each task whose inputs are the values true and false, and let a wrapper notebook decide which path to call","Add an If/else task with {{tasks.qc.values.bad_records}} > 0, with quarantine on true and publish on false","Add quarantine with a run-if condition of All failed and publish with All succeeded, both depending on qc","Pass bad_records as a job parameter to both notebooks and let each notebook exit early when it should not run"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/tasks/if-else","referenceTitle":"Add branching logic to a job with the If/else task — Databricks Docs"}}
{"id":"RAN-DBDEA-00062","exam":"DBDEA","topics":["jobs"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:bc238056331479d9d05cd8074416bda1769b50d25a49fc6484004779d9fddb49","aliases":[],"snapshot":{"statement":"A Lakeflow Job task named agg ends with:\n\n```python\ndbutils.jobs.taskValues.set(key = \"row_count\", value = result_count)\n```\n\nresult_count is the Python float `12.0`. An If/else condition task evaluates `{{tasks.agg.values.row_count}} == 12`, and the engineer expects the true branch to run — but the condition evaluates to false.\n\nWhy did this happen, and which comparison behaves numerically?","options":["Task values lose precision when serialized, so == is unreliable; >= has the same precision problem","If/else conditions accept only boolean task values, so numeric conditions silently evaluate to false","== and != compare as strings, so 12.0 does not equal 12; >, >=, <, and <= are numeric","The If/else task failed to read the value, because task values cannot store numbers"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/tasks/if-else","referenceTitle":"Add branching logic to a job with the If/else task — Databricks Docs"}}
{"id":"RAN-DBDEA-00063","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:46e70cdb7d1f7d6d34691354edb40c1add778ad9e181e6c82af1cc8278410d67","aliases":[],"snapshot":{"statement":"A team maintains a cleaning notebook that takes a table name as a parameter. They want one Lakeflow Job to run this notebook once for each of six bronze tables — sales, orders, returns, customers, products, and suppliers — with at most two iterations running at the same time.\n\nWhich approach should they use?","options":["A For each task with a JSON array of the table names as Inputs, a nested notebook using {{input.table}}, and Concurrency 2","A For each task containing six nested For each tasks, one per table, all sharing a single concurrency setting that the team tunes per run","Six If/else condition tasks, one per table, each branching to the cleaning notebook","One notebook task that loops over the six table names in Python and applies the cleaning logic sequentially"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/tasks/for-each","referenceTitle":"Use a For each task to run another task in a loop — Databricks Docs"}}
{"id":"RAN-DBDEA-00064","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:74eb07f5e0c1405c42ac1c785a086abb4d31a742e3b34edca01c815790ad2bf1","aliases":[],"snapshot":{"statement":"The first task in a Lakeflow Job, named load_raw, ends with:\n\n```python\ndbutils.jobs.taskValues.set(key = \"ingested_rows\", value = raw_df.count())\n```\n\nThe next task, validate_counts, must use that number as the expected row count in its own parameters.\n\nHow should validate_counts receive the value?","options":["Task values become global variables that every subsequent task in the job can read directly","Call dbutils.jobs.taskValues.set() again inside validate_counts, using the same key","Read the value with spark.conf.get('ingested_rows') in the downstream task","Reference it in the task parameters with the dynamic value {{tasks.load_raw.values.ingested_rows}}"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/task-values","referenceTitle":"Use task values to pass information between tasks — Databricks Docs"}}
{"id":"RAN-DBDEA-00065","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:25a8e5824023c4aebe645064a46086c8910ece5d313280ea60c5d5a1cfd73fd6","aliases":[],"snapshot":{"statement":"A notebook task in a Lakeflow Job must receive a processing date and an environment name. The notebook already starts with:\n\n```python\nbatch_date = dbutils.widgets.get(\"batch_date\")\nenv = dbutils.widgets.get(\"env\")\n```\n\nHow should the data engineer configure the task so the notebook receives these values?","options":["Declare the two values with dbutils.jobs.taskValues.set() in a task that runs before the notebook","Add batch_date and env as key-value Parameters on the notebook task","Store the values as environment variables on the job cluster and read them with os.environ","Set the values as Spark configurations in the task's advanced options"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/tasks/notebook","referenceTitle":"Notebook task for jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00066","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:4700e7bcb64acf1a28853587873ecdbee1c04d4d5a6370e0274414f5ea130385","aliases":[],"snapshot":{"statement":"A team is building a Lakeflow Job with three steps: (1) run a saved Databricks SQL query that aggregates yesterday's orders into a reporting table; (2) run an update of an existing Lakeflow Spark Declarative Pipelines (formerly Delta Live Tables) pipeline that rebuilds the silver layer; (3) refresh a published dashboard so BI users see updated results.\n\nWhich task types should they configure for these steps?","options":["A SQL task, a Pipeline task, and a Dashboard task","Three notebook tasks that call spark.sql(), dbutils, and displayHTML() respectively","A single For each task that loops over the three assets","Three Run Job tasks, one per asset, chained by dependencies"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/configure-task","referenceTitle":"Configure and edit tasks in Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00067","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0b20e5bbcf732e83b94b4017e616647ffb1b2f7da158c120768c396c79fd1ff7","aliases":[],"snapshot":{"statement":"A Lakeflow Job includes a SQL task that runs a saved query against Unity Catalog tables, alongside notebook tasks that run on a classic job cluster. Before approving the job, the workspace admin wants to confirm exactly which compute resource the SQL task will use.\n\nWhich statement is correct?","options":["The SQL task runs on the job's classic job cluster if one is configured, and fails otherwise when no cluster is attached","The SQL task starts its own Lakeflow pipeline to execute the query","The SQL task runs against a SQL warehouse — serverless is recommended, and pro is also supported","The SQL task runs on whatever all-purpose cluster is currently attached to the SQL editor"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/compute","referenceTitle":"Compute for Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00068","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:485999a5dcc33308ad614419a9c5f270c283e81a8a60bf510c8d83731e705420","aliases":[],"snapshot":{"statement":"A Lakeflow Job on classic compute has four tasks that run sequentially, and each task is configured with its own job cluster. Every task spends several minutes waiting for its cluster to start before doing any work. The team wants to reduce this startup overhead within each run.\n\nWhat should they do?","options":["Have the four tasks share a single jobs compute resource for the whole run","Replace the job clusters with one all-purpose cluster that stays running between job runs","Add artificial dependencies so that all four clusters start in parallel before the tasks run","Increase the worker count on each job cluster so that cluster startup finishes sooner"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/compute","referenceTitle":"Compute for Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00069","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:5e47846349ce2baadaa86ad8b34eb732ba3ed7e49132e1f373d0f23252ff5aa0","aliases":[],"snapshot":{"statement":"A Lakeflow Job runs every 30 minutes on a schedule, and each run takes about 40 minutes. The team notices that some scheduled runs never execute: while a previous run is still active, the new run appears as skipped. The job was created with the Jobs REST API, and queueing was not enabled.\n\nWhat is the cause, and how can the team fix it?","options":["The scheduler enforces a 10-second minimum interval between runs and drops anything closer; the schedule must be shortened","Maximum concurrent runs defaults to 1 and extra runs are skipped — raise it, or enable queueing","Scheduled jobs cannot overlap by design; convert the trigger to a continuous trigger to serialize the runs","Skipped runs indicate an invalid cron expression; the schedule must be rewritten in Quartz cron syntax"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00070","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e887c70451c00c54028a907f267011086feb1b1c6a657490718ecc7c147ded49","aliases":[],"snapshot":{"statement":"A vendor drops CSV files into an Amazon S3 bucket at unpredictable times — sometimes several times a day, sometimes not for days. Today a cron job runs every 30 minutes and usually finds nothing new. The workspace has Unity Catalog enabled, and the bucket is registered as a Unity Catalog external location.\n\nWhich change best fits this data availability pattern?","options":["Keep the scheduled trigger but move it to every minute, to shrink the window for missed files","Switch to a continuous trigger so the job is always running and picks up files immediately","Replace the scheduled trigger with a file arrival trigger on the external location path","Use a table update trigger on the Delta table that the job writes"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/file-arrival-triggers","referenceTitle":"Trigger jobs when new files arrive — Databricks Docs"}}
{"id":"RAN-DBDEA-00071","exam":"DBDEA","topics":["jobs"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:28f1acfa4574701ecbae53307ef52d56344e717d55625def1f80089288bc77cd","aliases":[],"snapshot":{"statement":"A file arrival trigger monitors a landing directory in a Unity Catalog volume. Files arrive in bursts: each batch of 20-30 files lands within about a minute, and batches are separated by hours. The team wants every complete batch processed by a single job run, and no more than one run every 15 minutes.\n\nWhich advanced options should they set on the trigger?","options":["Minimum time between triggers = 900 seconds only — runs are rate-limited, but a run can start as soon as the first files of a batch are detected, before the batch finishes landing","Wait after last change = 60 seconds only — batches are debounced, but nothing caps the frequency when batches arrive close together","Minimum time between triggers = 0 and Wait after last change = 0 — Databricks batches arrivals automatically by default","Minimum time between triggers = 900 seconds and Wait after last change = 60 seconds"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/file-arrival-triggers","referenceTitle":"Trigger jobs when new files arrive — Databricks Docs"}}
{"id":"RAN-DBDEA-00072","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:45a9f3a0f91b144cdb812cbeb0505aaf8fe2f61dd3d87419248c04308ce8ab72","aliases":[],"snapshot":{"statement":"Two upstream teams independently update the Unity Catalog Delta tables sales.orders and sales.customers at unpredictable times. A downstream Lakeflow Job must run only after both tables have been updated since the previous run. The team wants to avoid an always-on cluster and does not control how the upstream teams write their data.\n\nWhich trigger should the job use?","options":["A scheduled trigger that runs every 10 minutes and checks both tables for new commits in a notebook","A table update trigger monitoring both tables, configured to run when all monitored tables are updated","A file arrival trigger on the cloud storage paths that back both tables","Two jobs with a table update trigger on each table, chained by a Run Job task"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/trigger-table-update","referenceTitle":"Trigger jobs when source tables are updated — Databricks Docs"}}
{"id":"RAN-DBDEA-00073","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:a2e455db70c77118c8fb4b88b976ed4712f3766970377ea07d5e88d85eeb96ab","aliases":[],"snapshot":{"statement":"A compliance report must be generated every day at 06:00 UTC, using whatever data exists at that moment — even if no new data arrived overnight — and it must never run more than once per day. Data arrival times vary widely.\n\nWhich trigger type should this job use?","options":["A file arrival trigger on the landing zone, because data-driven triggers are always preferable to time-based ones","A table update trigger on the report's source tables, with a 24-hour minimum time between triggers","A continuous trigger that regenerates the report whenever the previous run finishes","A scheduled trigger with a cron expression that fires daily at 06:00 UTC"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/triggers","referenceTitle":"Automate jobs with schedules and triggers — Databricks Docs"}}
{"id":"RAN-DBDEA-00074","exam":"DBDEA","topics":["jobs"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:1ec5df533d5240feda52bef7e08d66bd649787697654c99e381cb5a99284aeee","aliases":[],"snapshot":{"statement":"A Lakeflow Job has task retries configured. The on-call data engineer notices that when a task fails on its first attempt but succeeds on a retry, no notification is ever sent — the team learns about these incidents later, from downstream consumers. They want to be notified after every failed task attempt, including attempts that will be retried.\n\nWhat should they change?","options":["Add another job-level failure notification with a different email address","Switch the notification destination from email to a webhook, which bypasses retry suppression and always fires immediately","Add failure notifications on the task itself — job-level ones stay silent during retries","Enable Mute notifications until the last retry on the existing job-level notification"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/notifications","referenceTitle":"Add notifications on a job — Databricks Docs"}}
{"id":"RAN-DBDEA-00075","exam":"DBDEA","topics":["jobs"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:ffa2beab48c7142bdb064bcc1f5d8be3424f2be9579dbcc301fcd8ab1620e68c","aliases":[],"snapshot":{"statement":"A Lakeflow Job with eight tasks failed during last night's run: tasks 1-4 succeeded, task 5 failed because of a bad record that has now been fixed in the source, and tasks 6-8 were skipped. Re-running tasks 1-4 is expensive because they re-ingest a full day of data.\n\nWhat is the most efficient way to recover?","options":["Run now, which re-executes all eight tasks from the beginning","Repair run on the failed run, which re-runs only the unsuccessful task 5 and the dependent tasks 6-8","Repair run on the failed run, which re-runs only task 5 and leaves tasks 6-8 permanently skipped","Clone the job, delete tasks 1-4 from the clone, and run the clone manually"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/repair-job-failures","referenceTitle":"Troubleshoot and repair job failures — Databricks Docs"}}
{"id":"RAN-DBDEA-00076","exam":"DBDEA","topics":["cicd"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:01587aa2dfc359aba5bea78a11969eec3b5faf0c0085bd54b23112530f92754a","aliases":[],"snapshot":{"statement":"A data engineer maintains the notebooks for a sales pipeline in a Databricks Git Folder (formerly Repos) connected to the team's GitHub repository. They need to add a new validation step to the `silver_orders` notebook, and team policy requires a peer review of every change before it lands on `main`.\n\nWhich workflow should the engineer follow entirely from the workspace UI?","options":["From the Git Folder: create a branch, edit and commit the notebook, push, then create a pull request against main","Edit the notebook directly on main, push the change immediately, and rely on the workspace file version history to roll back if the reviewer later objects","Export the notebook as a source file, edit it in a local IDE, and email the modified file to the reviewer before copying it back into the Git Folder","Clone the notebook into a personal workspace folder, collect review comments on the clone, and then overwrite the main notebook with the approved version"],"answer":"A","reference":"https://docs.databricks.com/aws/en/repos/git-operations-with-repos","referenceTitle":"Git operations with Git folders — Databricks Docs"}}
{"id":"RAN-DBDEA-00077","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0a904e02db962ebf88ba2c5ab6d458463826984fe459aff692535121c842d0f3","aliases":[],"snapshot":{"statement":"A data engineer is refactoring ingestion notebooks on the `feature/dq-checks` branch checked out in their Git Folder. While waiting for CI, they must also reproduce a bug that a teammate already fixed on the `hotfix/schema-fix` branch, running both branches' notebooks in the same workspace at the same time to compare outputs.\n\nWhat should the engineer do?","options":["Keep switching the existing Git Folder between feature/dq-checks and hotfix/schema-fix whenever they need to execute the other branch's code","Create an additional Git Folder for the same repo with hotfix/schema-fix checked out, so both branches coexist in the workspace","Run a %git checkout hotfix/schema-fix command as the first cell of the buggy notebook to load the teammate's version for that notebook only","Ask the teammate to merge the fix into feature/dq-checks so that both changes can be tested in a single branch"],"answer":"B","reference":"https://docs.databricks.com/aws/en/repos/git-folders-concepts","referenceTitle":"Git folders concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00078","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:8e22aa0e923df39bb019f9e66228d0c5ea32b93b03d787da0a728e35cd2422fc","aliases":[],"snapshot":{"statement":"A team stores all pipeline code in Git Folders backed by an Azure DevOps repository. Policy forbids direct pushes to `main`: every change must be reviewed in a pull request on the Git provider. A junior engineer has just finished work on the branch `feature/mart-fix` inside their Git Folder and must open the PR for the team's reviewers.\n\nWhich action meets the requirement directly from the workspace?","options":["Grant the reviewers Can Read permission on the Git Folder so they can open the branch and approve the changes there","Merge feature/mart-fix into main locally in the Git Folder, push the merge, and then create a PR to revert it if reviewers object","Push the branch from the Git Folder and create the PR from its pull request option — it opens on Azure DevOps","Export the branch as a DBC archive and attach it to a review ticket for the team's sign-off"],"answer":"C","reference":"https://docs.databricks.com/aws/en/repos/git-operations-with-repos","referenceTitle":"Git operations with Git folders — Databricks Docs"}}
{"id":"RAN-DBDEA-00079","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:5bdea4e07a4c043b555a1ecb50fa813a6962c3b8d32baa58beb88ddd8f1020ca","aliases":[],"snapshot":{"statement":"A team owns one ETL codebase that runs in dev, test, and prod workspaces, and each workspace uses a different SQL warehouse for the job's SQL tasks. The bundle's task references a variable wherever the warehouse is used:\n\n```yaml\nresources:\n  jobs:\n    nightly_orders:\n      tasks:\n        - task_key: load\n          sql_task:\n            warehouse_id: ${var.warehouse_id}\n```\n\nThe rest of the bundle defines the variable and the targets (the test target follows the same pattern as prod):\n\n```yaml\nvariables:\n  warehouse_id:\n    default: \"9d4e8f1a2b6c\"\n    description: SQL warehouse to use for SQL tasks\n\ntargets:\n  dev:\n    default: true\n  prod:\n    variables:\n      warehouse_id: \"e2f1c9d4a7b3\"\n```\n\nHow does this design serve three environments from a single codebase?","options":["Every databricks bundle deploy pauses and prompts the engineer to type the warehouse ID for the target being deployed","The variable is resolved when the job runs, reading the correct warehouse from each workspace's environment configuration","The team maintains three copies of this bundle, each hardcoding its own warehouse_id, and deploys the copy that matches the environment","Variables resolve per target at deploy time — dev uses the default, prod uses its override, resources stay identical"],"answer":"D","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/overrides","referenceTitle":"Overrides for Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00080","exam":"DBDEA","topics":["cicd"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:f90c50b492818c4a7d0a9004641722f5def83270caba1a61f5b99bedbc113651","aliases":[],"snapshot":{"statement":"A team's bundle defines a Lakeflow Spark Declarative Pipeline (formerly DLT) with one notebook library at the root, and the prod target overrides the `libraries` list:\n\n```yaml\nbundle:\n  name: etl_bundle\n\nresources:\n  pipelines:\n    sales_silver:\n      name: sales_silver\n      libraries:\n        - notebook:\n            path: ./src/dlt_silver.ipynb\n\ntargets:\n  prod:\n    resources:\n      pipelines:\n        sales_silver:\n          libraries:\n            - notebook:\n                path: ./prod/dlt_silver.ipynb\n```\n\nAfter running `databricks bundle deploy -t prod`, which notebook does the deployed pipeline use?","options":["Both ./src/dlt_silver.ipynb and ./prod/dlt_silver.ipynb, because target-level lists are merged with root-level lists","Only ./prod/dlt_silver.ipynb, because sequence (list) settings in a target replace the root-level list instead of merging with it","Only ./src/dlt_silver.ipynb, because root-level resource definitions always take precedence over target overrides of any resource setting","The deploy fails with a duplicate-resource error because sales_silver is defined in two places"],"answer":"B","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/overrides","referenceTitle":"Overrides for Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00081","exam":"DBDEA","topics":["cicd"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6d330042227c5407dee573530782c8784cab0f35fd0b0236e70da33a18b79fb2","aliases":[],"snapshot":{"statement":"A data engineering team supports dev, test, and prod workspaces. Today an administrator recreates the team's Lakeflow Job, its notebooks, and a Lakeflow Spark Declarative Pipeline in each workspace by clicking through the UI, and the definitions drift apart between environments. The team wants the environment definitions to live in Git and be deployed consistently.\n\nWhich Databricks capability is designed for this?","options":["Automation Bundles: YAML definitions of jobs, pipelines, and notebooks in Git, deployed per target via the CLI","Workspace export and import of DBC archives between workspaces, executed manually by an administrator for each release","A meta Lakeflow Job whose notebook tasks call the Jobs REST API to create the team's other jobs in each workspace","Copying the Delta table data between workspaces so that each environment rebuilds the pipeline from the same files"],"answer":"A","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00082","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:232d829ec4461bea6436128f5319d50b49bcc30b4ad2d7830d1181063d2de637","aliases":[],"snapshot":{"statement":"In a CI workflow, this command just completed with no errors reported:\n\n```bash\ndatabricks bundle validate -t prod\n```\n\nThe next step is to create or update the bundle's Lakeflow Job and pipeline in the production workspace so that they match the definitions in the repository.\n\nWhich command performs this step?","options":["databricks bundle push -t prod","databricks jobs create --json-file databricks.yml","databricks bundle deploy -t prod","databricks bundle run -t prod"],"answer":"C","reference":"https://docs.databricks.com/aws/en/dev-tools/cli","referenceTitle":"Databricks CLI — Databricks Docs"}}
{"id":"RAN-DBDEA-00083","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f63aca3c91c1ebe2fb049993e4a3a82e10457c1bcc2165d582a8f865e7d78f16","aliases":[],"snapshot":{"statement":"A GitHub Actions workflow runs on every pull request that touches the team's Automation Bundle. The team wants the check to fail before any production resource changes when the bundle has schema errors, invalid resource fields, or variables that cannot be resolved for the prod target.\n\nWhich command should the workflow run?","options":["databricks bundle lint --target prod --output json","databricks bundle plan --target prod","yamllint databricks.yml","databricks bundle validate --target prod"],"answer":"D","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/jobs-tutorial","referenceTitle":"Tutorial: Deploy a Lakeflow Job with an Automation Bundle — Databricks Docs"}}
{"id":"RAN-DBDEA-00084","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e4dc20e4bf5f9956ab174d49891edf9ef51ab1ae784e3e48eafa98a836fd51fd","aliases":[],"snapshot":{"statement":"A nightly GitHub Actions workflow deploys the team's Automation Bundle to production. The workflow must authenticate to the Databricks workspace without any human interaction, and the security team has banned personal access tokens that belong to individual engineers in CI systems.\n\nWhich setup should the team use for the Databricks CLI in the workflow?","options":["Embed the credentials of a shared admin user in the workflow definition and run databricks auth login interactively at the start of every run","Create a service principal, then set DATABRICKS_HOST, DATABRICKS_CLIENT_ID, and DATABRICKS_CLIENT_SECRET as environment variables","Rotate a personal access token borrowed from a different engineer each month and store it as the workflow's secret","Run the CLI from a VM in the same VPC as the workspace so that network proximity replaces authentication"],"answer":"B","reference":"https://docs.databricks.com/aws/en/dev-tools/cli","referenceTitle":"Databricks CLI — Databricks Docs"}}
{"id":"RAN-DBDEA-00085","exam":"DBDEA","topics":["cicd"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:eb99ef052238e80dd2feff54635a16f2647dec1f56c61156888bda05950c4f15","aliases":[],"snapshot":{"statement":"A CI workflow just deployed the team's bundle to the dev target. As a smoke test, the workflow must start the deployed Lakeflow Job whose bundle resource key is `ingest_orders`, capture its outcome, and fail the workflow if the run errors.\n\nWhich command starts the job from the bundle?","options":["databricks bundle run ingest_orders --target dev","databricks bundle trigger ingest_orders --target dev","databricks jobs run-now --job-name ingest_orders","databricks bundle validate ingest_orders --target dev --execute"],"answer":"A","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/jobs-tutorial","referenceTitle":"Tutorial: Deploy a Lakeflow Job with an Automation Bundle — Databricks Docs"}}
{"id":"RAN-DBDEA-00086","exam":"DBDEA","topics":["cicd"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:14a039d1341e433276000bb602e2cc9b7a59a0f034d12409dd4fcd4f9a154bf9","aliases":[],"snapshot":{"statement":"A data engineering team's bundle stores the Unity Catalog location for its pipeline in a variable, and the pipeline resource references that variable. The prod target assigns the variable its own value:\n\n```yaml\nbundle:\n  name: finance_etl\n\nvariables:\n  catalog:\n    default: dev_main\n\ntargets:\n  prod:\n    variables:\n      catalog: prod_main\n```\n\n```yaml\nresources:\n  pipelines:\n    finance_silver:\n      catalog: ${var.catalog}\n```\n\nThe engineer validates the bundle, then runs `databricks bundle deploy -t prod` from the terminal. In which catalog do the deployed pipeline's tables land?","options":["In dev_main, because a variable's default value always wins over target-level values","The deploy pauses and prompts the engineer to choose the catalog interactively","In prod_main, because the prod target's variable value overrides the default when deploying that target","Nowhere — the deploy fails because catalog is declared both as a default and in the prod target"],"answer":"C","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00087","exam":"DBDEA","topics":["cicd"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:a25216f774e1a924c104352f360e4443570557b1bfd83cbdd6b0fc1dca9dd91b","aliases":[],"snapshot":{"statement":"A team's `prod` target declares no `mode`. After deploying it, the job appears in the Jobs UI with a `[dev]` marker appended to its name, and the job runs under the identity of whichever engineer executed the deploy. The team wants production resources with exact names running under a stable identity.\n\nWhat should they change in the bundle?","options":["Set mode: production on the dev target and remove the dev target's default: true","Keep the bundle as is and rename the job in the Jobs UI after every production deploy","Split the job into a separate bundle that only the dev team deploys, leaving the current bundle for production","Set mode: production and run_as (service principal) on the prod target — exact names, stable identity"],"answer":"D","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/deployment-modes","referenceTitle":"Automation Bundle deployment modes — Databricks Docs"}}
{"id":"RAN-DBDEA-00088","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:472bea3d608993580c3dc0f3cc82c47e737cab96437a4295def0a2acac39ece0","aliases":[],"snapshot":{"statement":"A scheduled Lakeflow Job that loads nightly sales data used to complete in about 18 minutes. Over the past month, runs have crept up to 45 minutes and more. Before investigating causes, the on-call engineer wants to confirm whether this is a steady trend or a one-off spike.\n\nWhere should the engineer look first, and what will it show?","options":["The job's run history in the Jobs UI, comparing recent run durations against the earlier baseline","The Ganglia metrics of the job cluster, aggregated across all jobs that shared that cluster over the past month","The query history of the workspace's SQL warehouses, filtered by the job owner","The workspace audit logs, filtered by the job's service principal"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/diagnose-job-performance","referenceTitle":"Diagnose job performance — Databricks Docs"}}
{"id":"RAN-DBDEA-00089","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9b9926b841de8f4100efd137de4127a08127cb9d3cebdac47c4646d650c18e60","aliases":[],"snapshot":{"statement":"A Lakeflow Job has the task graph ingest → validate → enrich → publish. In the latest run, `ingest` succeeded, `validate` failed after four minutes, and `enrich` and `publish` never started — the task graph shows them as skipped.\n\nWhat does the task graph reveal, and what is the immediate cause?","options":["enrich is misconfigured: its skipped status means its own task parameters failed validation before launch","Failed validate blocks the rest: enrich and publish wait on upstream success and were skipped","The job ran out of compute: the scheduler skipped the tasks because no workers were available to launch them","enrich and publish were skipped because they were disabled in the job definition during the last edit"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/monitor","referenceTitle":"Monitor Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00090","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d923c9ba17ea7f2e71eeb692b08a4c99c16b3bb3511f991ffa44454f001cb854","aliases":[],"snapshot":{"statement":"A nightly Lakeflow Job with six tasks failed at `transform_customers` (task 4) because a partner delivered a malformed file. Tasks 1-3 succeeded, and tasks 5-6 were skipped. The partner has now re-delivered a corrected file, and the team wants tonight's data finished as quickly as possible without recomputing work that already succeeded.\n\nWhat should the team do?","options":["Trigger a brand-new run of the job and let it recompute tasks 1 through 6 from scratch","Run the transform_customers notebook manually on an all-purpose cluster, copy the resulting tables into place, and then trigger a new run","Re-run the failed run with repair — it re-runs the failed task and its dependents, preserving succeeded tasks","Delete the failed run so that the job's next scheduled trigger retries the full pipeline automatically"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/repair-job-failures","referenceTitle":"Repair Lakeflow Job run failures — Databricks Docs"}}
{"id":"RAN-DBDEA-00091","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:7e0889a233c0ea8512a6bba102506198def2d9d735ed37fbd73ebf9c6d516ba6","aliases":[],"snapshot":{"statement":"A PySpark job joins a 4 TB events table with a small dimension table on `user_id`. One stage has 200 tasks: 199 finish in under two minutes, but one task runs for 38 minutes. The stage-level summary metrics show:\n\n```text\nShuffle read size per task:  min 14 MB  |  median 32 MB  |  max 7.9 GB\nTask duration:               min 21 s   |  median 48 s   |  max 38 min\n```\n\nWhat is the most likely diagnosis?","options":["Too few shuffle partitions: all tasks are uniformly too large, so spark.sql.shuffle.partitions should be increased to shrink every task","Disk spill: the stage is dominated by spilling shuffle data to local disk, which the duration column reveals","A broadcast timeout: the small table exceeded the broadcast threshold and the join fell back to a shuffle","Data skew on user_id: one partition received 7.9 GB of shuffle data versus a 32 MB median, producing a single long-running straggler task"],"answer":"D","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/long-spark-stage","referenceTitle":"Investigate long-running stages in the Spark UI — Databricks Docs"}}
{"id":"RAN-DBDEA-00092","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:2577f9b64c7a6d6de956ec73cbcc68633f90d8d37bdeef23655c6d25a0aeb0d3","aliases":[],"snapshot":{"statement":"A pipeline joins a large fact table with a dimension table on a heavily skewed key. The team cannot rewrite the join right now and wants Spark to mitigate the skew automatically at runtime.\n\nWhat does Adaptive Query Execution (AQE) provide for this case?","options":["AQE converts every skewed sort-merge join into a broadcast join using the smaller table's runtime statistics","AQE's skew join optimization splits oversized partitions into smaller ones processed in parallel","AQE caches the skewed partitions in memory on the executors so that repeated reads avoid the shuffle","AQE disables the join and rewrites it as a union of per-key filters generated from table statistics"],"answer":"B","reference":"https://docs.databricks.com/aws/en/optimizations/aqe","referenceTitle":"Adaptive Query Execution — Databricks Docs"}}
{"id":"RAN-DBDEA-00093","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:71ad07ec663b7b42de6a0e3e75944bbe9be71aba05b618adb8e70689ccdf9aaa","aliases":[],"snapshot":{"statement":"In the Spark UI, a stage's tasks report large Spill (Disk) values — tens of GB spilled across the stage — and the stage takes three times longer than its siblings. Executor garbage-collection time is elevated, but no single task is disproportionately larger than the others.\n\nWhat does the spill indicate, and what is a reasonable remedy?","options":["The local SSDs are saturating; replace instance storage with network-attached volumes that can grow beyond the node's own disks","The Delta table has too many small files; run OPTIMIZE to compact them before the next run","Shuffle data per task exceeds executor memory, so it spills to disk — raise memory per executor or shuffle partitions","Spill means the DBIO disk cache is active; disable the cache to free disk space for shuffle files"],"answer":"C","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/","referenceTitle":"Spark UI guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00094","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:dc431f1ab09a362bf00d876e7e9a855ff499aafc5e807d3cbf6ec7c3d8b051cb","aliases":[],"snapshot":{"statement":"A team is creating a new Delta table for order line items. Queries almost always filter on `order_date` and `customer_id`. The team's older tables use Hive-style partitioning on a date column plus nightly `OPTIMIZE ... ZORDER BY` jobs, and the team plans to copy that design.\n\nWhat does Databricks currently recommend for this new table?","options":["Liquid Clustering with order_date and customer_id as keys — incremental layout, no ZORDER jobs","Hive-style partitioning on customer_id, because clustering keys must have low cardinality to be effective","Hive-style partitioning on both order_date and customer_id, so every query prunes at least one partition directory","No layout optimization at all: rely on Delta's file-level statistics, which are recomputed on every write"],"answer":"A","reference":"https://docs.databricks.com/aws/en/tables/clustering","referenceTitle":"Use Liquid Clustering for tables — Databricks Docs"}}
{"id":"RAN-DBDEA-00095","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:cf9c41f0daac306d845e5924b0578c8b17b2161b7f32b9d72b5eab4a1a5b8cf1","aliases":[],"snapshot":{"statement":"A Delta table was created with Liquid Clustering using `CLUSTER BY (region, order_date)`. Usage has shifted: most queries now filter by `customer_id`, and region filters are rare. The table holds several years of data that must not be rewritten wholesale.\n\nHow can the clustering keys be adapted?","options":["They cannot be changed: clustering keys are fixed at table creation, so the team must copy the data into a new table clustered by the new keys","Run OPTIMIZE orders ZORDER BY (customer_id) to layer ZORDER columns on top of the existing clustering","Recreate the table with CREATE OR REPLACE TABLE orders CLUSTER BY (region, customer_id) and reinsert all rows during the next maintenance window","Run ALTER TABLE orders CLUSTER BY (region, customer_id): new keys apply to new writes, and OPTIMIZE re-clusters incrementally"],"answer":"D","reference":"https://docs.databricks.com/aws/en/tables/clustering","referenceTitle":"Use Liquid Clustering for tables — Databricks Docs"}}
{"id":"RAN-DBDEA-00096","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:bb531faabf419fe0f43f2ce92ddbea7f69e8eec78a4354ac1eb6e7142276c2b0","aliases":[],"snapshot":{"statement":"A small platform team schedules dozens of manual Lakeflow Jobs to run OPTIMIZE and VACUUM on Unity Catalog managed Delta tables. Runs get skipped whenever the on-call rotation changes, and table performance decays unpredictably. They have heard that Databricks can handle this maintenance automatically.\n\nWhat does predictive optimization provide?","options":["It forecasts query workloads and pre-materializes the most frequent gold-layer joins as streaming tables","Databricks runs OPTIMIZE and VACUUM automatically on UC managed tables based on usage patterns","It predicts cluster node failures and pre-emptively migrates jobs to healthy workers","It analyzes table growth trends and emails capacity recommendations to the data engineering team"],"answer":"B","reference":"https://docs.databricks.com/aws/en/optimizations/predictive-optimization","referenceTitle":"Predictive optimization — Databricks Docs"}}
{"id":"RAN-DBDEA-00097","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:986e55ecca25d4c2d26663b185d85527fe843d7bb1e6251d848c9d03348cee18","aliases":[],"snapshot":{"statement":"A shuffle-heavy aggregation job fails repeatedly with `java.lang.OutOfMemoryError: Java heap space` on the executors. The Spark UI shows long garbage-collection pauses and heavy spill on the failing stage, and the cluster uses the smallest general-purpose worker type, with each node running multiple concurrent tasks.\n\nWhat is the most effective fix?","options":["Increase driver memory, because the driver's heap backs every executor task on the cluster","Decrease spark.sql.shuffle.partitions so that fewer tasks run concurrently and the total memory demand of the stage drops accordingly","Bigger workers (more memory per executor) plus more shuffle partitions, so each task handles less data","Disable the Spark UI for this job to release the executor heap consumed by the event timeline"],"answer":"C","reference":"https://docs.databricks.com/aws/en/compute/cluster-config-best-practices","referenceTitle":"Cluster configuration best practices — Databricks Docs"}}
{"id":"RAN-DBDEA-00098","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ad656c7951500e6e8953682397ea44d0666d3c417c22664fba65267cac34e830","aliases":[],"snapshot":{"statement":"After a platform change, every new job cluster terminates during launch. The cluster event log shows an `INIT_SCRIPTS_FAILED` event, and the team recently added an init script that points pip at an internal package mirror whose TLS certificate expired last night.\n\nWhat is the cause, and how is it fixed?","options":["The init script fails during node startup, so Databricks terminates the cluster before Spark launches","The Databricks Runtime version conflicts with the cluster's Spark settings; pin the previous version","Executors run out of memory at boot because the init script pre-loads libraries into the heap","The cloud provider has no capacity for the instance type; launch the cluster in a different availability zone"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00099","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:f8c91052c29df962801457d26b66f8522645eaeed154c7345a972d2c68f17a63","aliases":[],"snapshot":{"statement":"A job ran unchanged for months on a shared cluster whose `pandas` cluster library was pinned to 1.5.3. Last week, another team updated the cluster's libraries to pandas 2.2.0 for their own pipeline. Since then, this job's tasks fail immediately with an `AttributeError` when the code calls `DataFrame.append`.\n\nWhat happened, and what is the best-practice fix?","options":["The cluster's Spark version was silently downgraded during the library update; pin the previous Spark version to restore the job","The other team's pipeline rewrote the job's input files into a format the reader no longer understands","The job hits out-of-memory while importing pandas; move it to larger workers","A library version conflict — pandas 2.2.0 removed the API this job uses; give the job its own cluster or pinned libraries"],"answer":"D","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00100","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ad9498ecdb263389ff5482ab30f804f4b84a56215d383589f228b0e28f933c63","aliases":[],"snapshot":{"statement":"A data engineer is decommissioning two production tables governed by Unity Catalog. Both tables currently hold live data, and the engineer runs the following statements, which complete successfully:\n\n```sql\nDROP TABLE main.sales.orders;       -- originally created with CREATE TABLE main.sales.orders (...);\nDROP TABLE main.sales.orders_hist;  -- originally created with CREATE TABLE ... LOCATION 's3://acme-raw/orders_hist';\n```\n\nThe cloud storage team asks what will happen to the underlying data files once the DROP statements complete. Which answer is correct?","options":["orders (managed): UC deletes the data files; orders_hist (external): the files stay in cloud storage.","The files for both tables are deleted, because DROP TABLE in Unity Catalog always removes the underlying data.","The files for both tables remain in place, because DROP TABLE in Unity Catalog removes only table metadata.","The files for orders remain in place, while the files for orders_hist are deleted from the external location."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/managed-versus-external","referenceTitle":"Managed versus external assets in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00101","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ab0fdbc0c45b7c56a33ce59a2f31b7c19fb2801ddda29b542ac64ef69b676c2a","aliases":[],"snapshot":{"statement":"An upstream system continuously writes Delta files to s3://acme-raw/legacy/events/, and that system must keep full control of the data's lifecycle. The data team wants to query this existing data with SQL in Unity Catalog without copying or moving it.\n\nWhich statement registers the existing files as a Unity Catalog external table?","options":["CREATE TABLE main.legacy.events AS SELECT * FROM read_files('s3://acme-raw/legacy/events');","CREATE EXTERNAL TABLE main.legacy.events USING DELTA LOCATION 's3://acme-raw/legacy/events';","CREATE TABLE main.legacy.events MANAGED LOCATION 's3://acme-raw/legacy/events';","COPY INTO main.legacy.events FROM 's3://acme-raw/legacy/events' FORMAT AS DELTA;"],"answer":"B","reference":"https://docs.databricks.com/aws/en/tables/external","referenceTitle":"Work with external tables — Databricks Docs"}}
{"id":"RAN-DBDEA-00102","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ee3c9d3fb7e08723cdceafbbc84819aff16c8861224946e091a8c7ed9c3803d6","aliases":[],"snapshot":{"statement":"A senior engineer creates the following table in a schema that has a managed storage location configured:\n\n```sql\nCREATE TABLE main.fin.transactions (\n  txn_id BIGINT,\n  account_id STRING,\n  amount DECIMAL(12,2)\n);\n```\n\nDuring a compliance review, the auditor asks where this table's data files physically reside and who controls their lifecycle. Which answer is accurate?","options":["They are stored in storage operated by Databricks outside the company's cloud account, because Unity Catalog owns the data of managed tables.","They are stored in a path chosen by the table creator inside any cloud bucket the creator can access.","In the schema's managed storage location, inside the company's own cloud account — Unity Catalog controls the file lifecycle.","They are stored in the workspace's DBFS root directory, managed by workspace administrators."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/managed-versus-external","referenceTitle":"Managed versus external assets in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00103","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:389da7b70e5f8641aaee795df6fde4020a6ce118e5d8b063d06351f3df99714a","aliases":[],"snapshot":{"statement":"main.sales.orders_hist is a Unity Catalog external Delta table whose files live at s3://acme-raw/orders_hist. The team now wants Unity Catalog to manage the table's storage and lifecycle, keeping the same table name, permissions, and history, with minimal reader and writer downtime. The team owns the table and uses Serverless compute.\n\nWhich command should they run?","options":["ALTER TABLE main.sales.orders_hist SET MANAGED MOVE;","CREATE TABLE main.sales.orders_managed AS SELECT * FROM main.sales.orders_hist; followed by DROP TABLE main.sales.orders_hist;","ALTER TABLE main.sales.orders_hist UNSET MANAGED;","ALTER TABLE main.sales.orders_hist SET MANAGED;"],"answer":"D","reference":"https://docs.databricks.com/aws/en/tables/convert-to-managed","referenceTitle":"Convert external or foreign Delta Lake tables to Unity Catalog managed tables — Databricks Docs"}}
{"id":"RAN-DBDEA-00104","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:0abe0076a90aee7bb80f91ae88e5a40476095412a48573ed353f8e8be0b99ee2","aliases":[],"snapshot":{"statement":"A data team is building a new curated table that summarizes several silver tables, computed with a single query. They want Unity Catalog to own the table's storage location, optimization, and deletion — no external system will ever write to these files.\n\nWhich single statement creates the table with the intended ownership of the data lifecycle?","options":["CREATE TABLE gold.daily_summary LOCATION 's3://acme-gold/summary' AS SELECT ... FROM silver.orders;","CREATE TABLE gold.daily_summary AS SELECT ... FROM silver.orders;","CREATE VIEW gold.daily_summary AS SELECT ... FROM silver.orders;","df.write.option(\"path\", \"s3://acme-gold/summary\").saveAsTable(\"gold.daily_summary\")"],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/managed-versus-external","referenceTitle":"Managed versus external assets in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00105","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c961cd91925a28e06e297c254ad4ce3b847643c955a48eb84975666e172a6c6a","aliases":[],"snapshot":{"statement":"A lakehouse on Unity Catalog contains the catalog main with several schemas. The group data_analysts already has these grants:\n\n```sql\nGRANT USE CATALOG ON CATALOG main TO data_analysts;\nGRANT USE SCHEMA ON SCHEMA main.sales TO data_analysts;\n```\n\nAn ingestion job creates new tables in main.sales every week. The team wants analysts to read every current and future table in main.sales — and only in that schema — without maintaining grants each week. Which additional statement achieves this?","options":["GRANT SELECT ON SCHEMA main.sales TO data_analysts;","GRANT SELECT ON CATALOG main TO data_analysts;","GRANT SELECT ON METASTORE TO data_analysts;","GRANT SELECT ON TABLE main.sales.* TO data_analysts;"],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00106","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7bd91218f6753ea52c4e87c586825f8900199a8e124f249918f3cd452cfe1ca1","aliases":[],"snapshot":{"statement":"An analyst who belongs to the group finance_analysts runs:\n\n```sql\nSELECT * FROM analytics.finance.invoices;\n```\n\nThe query fails with a permissions error, even though the group's grants include:\n\n```sql\nGRANT SELECT ON TABLE analytics.finance.invoices TO finance_analysts;\n```\n\nThe group has SELECT on the table, yet the read still fails. Which additional privileges does the group need?","options":["MODIFY on analytics.finance.invoices, because reads also require the write privilege on Unity Catalog tables.","BROWSE on catalog analytics, which lets principals read data from all objects in the catalog.","USE CATALOG on analytics and USE SCHEMA on analytics.finance, in addition to SELECT on the table.","Ownership of analytics.finance.invoices must be transferred to finance_analysts before any member can query it."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00107","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:a190db60cf49823bae5772492bb047af648f992511ef39f425353cf437b41187","aliases":[],"snapshot":{"statement":"A governance engineer is onboarding a new analytics team to the lakehouse. The team's group, finance_analysts, must be allowed to read rows from the table analytics.finance.invoices, and the engineer wants to apply the grant with SQL in a notebook rather than through Catalog Explorer.\n\nWhich statement uses the correct Databricks SQL syntax?","options":["GRANT SELECT TO finance_analysts ON TABLE analytics.finance.invoices;","GRANT READ ON TABLE analytics.finance.invoices TO finance_analysts;","GRANT SELECT ON TABLE analytics.finance.invoices FROM finance_analysts;","GRANT SELECT ON TABLE analytics.finance.invoices TO finance_analysts;"],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00108","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:edb6bbf3068b97d80923665840abc3b7828d50dc1a3477e9d6d77855aa0365b4","aliases":[],"snapshot":{"statement":"A platform admin runs the following statements against a Unity Catalog metastore, intending to cut off a contractor's read access:\n\n```sql\nGRANT SELECT ON TABLE main.sales.orders TO data_analysts;\nREVOKE SELECT ON TABLE main.sales.orders FROM `contractor@acme.com`;\n```\n\nThe user contractor@acme.com is a member of data_analysts. A few minutes after the REVOKE completes, the contractor runs SELECT COUNT(*) FROM main.sales.orders and it succeeds. Why does the contractor keep access?","options":["REVOKE statements only take effect after the principal's active sessions end, and the contractor still has an open session in the workspace.","The contractor's SELECT comes from the group grant — REVOKE only removes direct grants, leaving group access intact.","REVOKE must be executed by a metastore admin to affect principals other than the statement's author.","The contractor also needs USE SCHEMA on main.sales before a table-level REVOKE can apply to them."],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00109","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:568dc0b7b6ba1d92936db663253140de6742c8e4941a6d94ed4262eafb50fe5e","aliases":[],"snapshot":{"statement":"A contractor's engagement is being transferred to a partner organization. Today the contractor reads main.hr.compensation through membership in the hr_analysts group, which holds SELECT on that table. Security policy requires that this specific principal must never be able to read the table again — even if the contractor is later added to other groups that hold SELECT on it, now or in the future.\n\nWhich action guarantees this outcome?","options":["Explicitly deny SELECT on the table to the contractor — it overrides any grant, group or direct.","Run REVOKE SELECT ON TABLE main.hr.compensation FROM contractor@acme.com.","Remove the contractor from the hr_analysts group.","Transfer ownership of main.hr.compensation to the security team and remove all existing grants on the table for every principal."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00110","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:2f69ff4cb82464a72879c9091cb0886f3a0354fc4807f2c5fa181800f1128401","aliases":[],"snapshot":{"statement":"A platform team supports about 60 analysts who join and leave the company frequently, plus a scheduled ingestion pipeline that runs unattended every night. The team wants to manage Unity Catalog access with minimal maintenance while keeping a clear audit trail of who did what.\n\nHow should they assign privileges?","options":["Grant SELECT on the analysts' tables to each analyst's individual user account, re-running the grants whenever someone joins or leaves.","Grant all privileges to a shared analytics_admin user account and have all analysts sign in with it.","Grant privileges to the nightly pipeline's cluster configuration, because compute-level grants cover everyone who attaches to it.","Add analysts to groups and grant privileges to the groups; run the pipeline under a service principal granted only the privileges it needs."],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00111","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:5cb90c17357583b2d242f119b9edac5c7dd14825a16da278405011a13cb138a4","aliases":[],"snapshot":{"statement":"A retail company stores all of its sales transactions in main.sales.transactions, with records from every sales region. The governance team protects the table with the following row filter:\n\n```sql\nCREATE FUNCTION main.sec.emea_filter(region STRING)\nRETURN IF(is_account_group_member('sales_admins'), true, region = 'EMEA');\n\nALTER TABLE main.sales.transactions SET ROW FILTER main.sec.emea_filter ON (region);\n```\n\nMembers of sales_admins and analysts who are not members of the group both query main.sales.transactions. What does each user see?","options":["All users see all rows; the function only records which non-admin users queried the table.","The query fails for users outside sales_admins, because they lack permission to evaluate the filter function.","Members of sales_admins see all rows; everyone else sees only the rows where region = 'EMEA'.","All users see only EMEA rows in SQL, but members of sales_admins can bypass the filter by exporting the table from Catalog Explorer."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/filters-and-masks/","referenceTitle":"Row filters and column masks — Databricks Docs"}}
{"id":"RAN-DBDEA-00112","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ee3e545ac0641bb063769b357d670903a1d07ba55d1fd2a4294acb6b08ca8f3e","aliases":[],"snapshot":{"statement":"The table main.crm.customers contains the column email (STRING). Only the group pii_viewers should see real email addresses; every other user with SELECT should see ***@***.*** instead. A SQL UDF implements the logic:\n\n```sql\nCREATE FUNCTION main.sec.mask_email(email STRING)\nRETURN CASE WHEN is_account_group_member('pii_viewers')\n            THEN email\n            ELSE '***@***.***' END;\n```\n\nWhich statement correctly applies this function as a column mask on that column?","options":["ALTER TABLE main.crm.customers SET MASK main.sec.mask_email ON (email);","ALTER TABLE main.crm.customers ALTER COLUMN email SET MASK main.sec.mask_email;","ALTER TABLE main.crm.customers SET ROW FILTER main.sec.mask_email ON (email);","ALTER TABLE main.crm.customers ALTER COLUMN email SET ROW FILTER main.sec.mask_email;"],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/filters-and-masks/","referenceTitle":"Row filters and column masks — Databricks Docs"}}
{"id":"RAN-DBDEA-00113","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f7b772766077ebd06d5cec707e9359dec34624437451b0a7552431dfdfa73afd","aliases":[],"snapshot":{"statement":"An HR database contains employee records with sensitive Social Security Numbers. The ssn column of main.hr.employees is protected as follows:\n\n```sql\nCREATE FUNCTION main.sec.ssn_mask(ssn STRING)\nRETURN CASE WHEN is_account_group_member('hr_payroll') THEN ssn\n            ELSE '***-**-' || right(ssn, 4) END;\n\nALTER TABLE main.hr.employees ALTER COLUMN ssn SET MASK main.sec.ssn_mask;\n```\n\nAn HR analyst (member of hr_payroll) and a finance analyst (not a member) both hold SELECT on the table and run:\n\n```sql\nSELECT name, ssn FROM main.hr.employees;\n```\n\nWhat does each analyst see in the ssn column?","options":["Both see the full SSN values, because both principals hold SELECT on the table and masks only apply to writes.","Both see masked values; the raw values can only be read from the underlying files.","The finance analyst's query fails, because the mask blocks users outside hr_payroll.","The HR analyst sees the full SSN values; the finance analyst sees values such as ***-**-1234."],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/filters-and-masks/","referenceTitle":"Row filters and column masks — Databricks Docs"}}
{"id":"RAN-DBDEA-00114","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:4b900d40d2531984ae0c45314f98701589b071eaafda9dae5c36276badfbb356","aliases":[],"snapshot":{"statement":"A governance admin wants a single policy attached to the catalog main that masks every column tagged pii = ssn for all users except the compliance_team group. Hundreds of tables live under main, and new columns are tagged every week — they must be covered without editing the policy.\n\nIn Unity Catalog ABAC, what mechanism identifies which columns the policy protects?","options":["Governed tags: conditions such as pii = ssn, evaluated with has_tag_value(), match every tagged column — even ones tagged later.","Content inspection: Unity Catalog scans column values at query time and masks anything matching an SSN pattern, with no policy configuration.","Name heuristics: the policy infers sensitive columns from names such as ssn, tax_id, or national_id.","Manual registration: each table owner must list their table's columns in the policy."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/abac/","referenceTitle":"Attribute-based access control in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00115","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f37c76649326ec0db24e54cfb1bd24ea0a04671fa1d94fdc81cd1fcb8401ad7e","aliases":[],"snapshot":{"statement":"A bank runs 200+ Delta tables across the catalogs main and finance. Sensitive columns are classified with governed tags. The governance team must guarantee consistent masking rules for all of them, and — critically — table owners must not be able to weaken or remove the protections on their own tables. New tables are created daily and must be covered automatically once they are tagged.\n\nWhich approach meets these requirements?","options":["Each table owner writes a masking UDF and runs ALTER TABLE ... ALTER COLUMN ... SET MASK on their own tables.","ABAC column mask policies at the catalog level using governed tags: automatic coverage, and table owners cannot override them.","The governance team creates one dynamic view per table, embedding is_account_group_member() checks in each view definition.","The governance team grants SELECT on all sensitive tables only to the compliance_team group and tells everyone else to request extracts."],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/abac/abac-vs-rls-cm","referenceTitle":"When to use ABAC vs table-level row filters and column masks — Databricks Docs"}}
{"id":"RAN-DBDEA-00116","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:c8acfbaaa210fdfbf6366f670b3ec84c2d6af1838dbd2153aa9290cb77c02575","aliases":[],"snapshot":{"statement":"A lead data engineer is preparing a governance review for auditors who want to understand how Databricks protects sensitive data in Unity Catalog. The engineer must accurately compare Unity Catalog ABAC policies with table-level row filters and column masks.\n\nWhich statement about the two mechanisms is accurate?","options":["ABAC policies replace object-level grants: users matched by a policy can query protected tables without being granted SELECT.","Table-level row filters and column masks are evaluated by Unity Catalog before the query reaches the runtime, while ABAC policies are evaluated inside the query engine afterwards.","ABAC policies attach at catalog or schema level, match via governed tags, and cannot be modified by table owners; table-level masks are owner-managed per table.","Table-level column masks can be attached to many tables at once by tagging the tables, which makes them equivalent to ABAC policies."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/abac/abac-vs-rls-cm","referenceTitle":"When to use ABAC vs table-level row filters and column masks — Databricks Docs"}}
{"id":"RAN-DBDEA-00117","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:adbbb0e319631f8434769f7d95fd4a99916a634e6e97700bd58feccfe8b4fb20","aliases":[],"snapshot":{"statement":"A fintech runs its Databricks workspace on classic compute, with all-purpose clusters for interactive notebook work and job clusters for nightly Lakeflow Jobs. During a security review, the cloud architect asks in which cloud account those clusters will run, because that determines network routing and instance policies. Which statement correctly describes how compute is deployed for this workspace?","options":["Classic clusters run in the customer's cloud account, in the data plane; the Databricks-managed control plane hosts the UI, job scheduling, and APIs","Every deployment runs all of its compute in Databricks-owned cloud accounts, so customer cloud resources never execute any workloads","The data plane is only a metadata layer for the workspace UI, and all cluster workloads execute inside the Databricks control plane","Serverless compute runs inside the customer's VPC, while classic compute runs in Databricks-owned cloud accounts"],"answer":"A","reference":"https://docs.databricks.com/aws/en/introduction","referenceTitle":"Introduction to the Databricks Data Intelligence Platform"}}
{"id":"RAN-DBDEA-00118","exam":"DBDEA","topics":["platform"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:8a73749fc18d9e949f9832dd9174e499d760d11f5b9ce4ed65f53b77cea7aa1e","aliases":[],"snapshot":{"statement":"A data engineer created an all-purpose cluster for interactive notebook work on Friday, used it all day, and then forgot to terminate it before leaving. On Monday, the cloud bill shows the cluster ran — and billed — for the entire weekend, because an administrator had changed the cluster's termination settings at creation. Which default behavior would have stopped the cluster if those settings had been left unchanged?","options":["Interactive clusters never terminate on their own by default; termination must be configured and triggered manually","The same default inactivity timeout applies to job clusters, which otherwise stay running between their job runs","All-purpose clusters are created with auto-termination enabled by default, after 120 minutes of inactivity","Auto-termination only affects clusters that use spot instances; on-demand clusters keep running until stopped"],"answer":"C","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00119","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f8d7b38f380363774ac3e41abd00bebef1c5f355d68a0af89e5e3a649722eae3","aliases":[],"snapshot":{"statement":"A data engineering team is standardizing the environment for a long-lived production pipeline. The tech lead asks everyone to select a Databricks Runtime version for the pipeline's clusters, ideally one that will keep receiving fixes for as long as possible. Team members disagree about what the Databricks Runtime actually is and what an LTS version offers. Which statement is correct?","options":["Databricks Runtime is only the operating system image installed on the cluster's nodes, with no data processing libraries included","Databricks Runtime packages Apache Spark, Delta Lake, and core libraries, and LTS versions receive fixes for an extended period","Databricks Runtime LTS versions are the oldest and slowest releases available, so new pipelines should always use the latest non-LTS version instead","Databricks Runtime versions are picked automatically for each notebook session, and cannot be pinned on the cluster"],"answer":"B","reference":"https://docs.databricks.com/aws/en/compute","referenceTitle":"Compute — Databricks Docs"}}
{"id":"RAN-DBDEA-00120","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:742f58e8a9374299755135870d0e9911e1eeac90077736cbe6f538f8cd2e13cf","aliases":[],"snapshot":{"statement":"A new Auto Loader stream must ingest events from a landing directory that already holds two years of history: about one million small JSON files produced by a decommissioned legacy platform. Going forward, only a few dozen new files arrive each day. The stream's very first micro-batch tried to absorb the entire backlog at once and the job ran out of memory. How should the team bound the initial load?","options":["Split the source directory into one prefix per year and run one Auto Loader stream per year, all in parallel","Set cloudFiles.maxFilesPerTrigger to a fixed cap, so each micro-batch ingests at most that many files and the backlog drains gradually","Turn cloudFiles.useNotifications off for the first run, because file notification mode is what forces huge micro-batches","Raise the worker count on the job cluster so the single first micro-batch has enough memory to process the whole backlog"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00121","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:74a68a3fed222947efc9857954dfe463cfb6b92335eb29ed3ab634e0c701c750","aliases":[],"snapshot":{"statement":"A team starts a new Auto Loader stream on a bucket that already contains three years of legacy extracts from a retired system. Those historical files will be loaded once by a separate backfill script; the stream itself must capture only the files that arrive from its first run onward. How should the stream be configured so the legacy extracts are never ingested?","options":["Delete the legacy extracts from the bucket before starting the stream, so the first run finds nothing old to pick up","Use COPY INTO for this table instead, because COPY INTO always skips files that already existed in the folder before its very first execution","Create the target table with a date filter on ingestion time, which Auto Loader honors when it discovers new files","Set cloudFiles.includeExistingFiles to false, so the first run ignores files already in the bucket and ingests only new arrivals"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00122","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f8542d934cd30fff90485387728581520e19f962fe7e85fd851db87874e78fd2","aliases":[],"snapshot":{"statement":"An online retailer's operational data lives in a MySQL database: product catalog, inventory, and orders. The analytics platform needs an initial copy of every table in Unity Catalog-governed tables, followed by continuous replication of inserts, updates, and row deletions as they happen, with no ingestion code for the data team to write or operate. Which approach meets these requirements?","options":["A Lakeflow Connect database connector: it takes an initial snapshot of each table and then streams ongoing changes, including deletes, into Unity Catalog-governed streaming tables","A scheduled COPY INTO statement that reads the MySQL tables directly and appends their new rows into a Delta table every night","A nightly Lakeflow Job that reads full table snapshots over JDBC and overwrites the target Delta tables with each night's copy","A Lakeflow Connect Kafka connector pointed at the MySQL server, so row changes flow from the database into a streaming table"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/lakeflow-connect/","referenceTitle":"Lakeflow Connect connector concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00123","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:80742ffee08b486bff1c3617ed4cc906beaba0f2650f933ceeb86e602a50a767","aliases":[],"snapshot":{"statement":"A team defines a streaming table in a Lakeflow Declarative Pipeline using the read_files table-valued function over a landing directory, and the table's refresh schedule runs every 15 minutes. Files land in that directory around the clock. A new engineer assumes each refresh re-reads the entire directory and worries about duplicate rows. What does each scheduled refresh actually process?","options":["Only the files that have arrived since the last successful refresh, because the streaming table keeps durable state tracking what it has already ingested","Every file in the directory, with duplicate rows removed afterwards by a deduplication step that the pipeline runs automatically","Nothing, until an external scheduler updates a control table listing which files the previous refresh already consumed","Only new files if the table was declared with AUTO CDC INTO; streaming tables fed by read_files always reload in full"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ldp/concepts/streaming-tables","referenceTitle":"Streaming tables (Lakeflow pipelines) — Databricks Docs"}}
{"id":"RAN-DBDEA-00124","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6b1936ade29c0960ac6ef65e47f9669967eb8711a6e502e6025744292bfeba50","aliases":[],"snapshot":{"statement":"A Lakeflow Job runs an Auto Loader stream on a job cluster and terminates the cluster after each run. The stream's checkpointLocation points at a scratch path on storage that the job wipes and recreates empty whenever a new cluster starts. Every morning, the target Delta table contains duplicate rows re-ingested from the full landing directory. Where is the configuration error?","options":["The checkpointLocation points at ephemeral storage; the checkpoint must live in durable cloud storage, in a location dedicated to this stream alone","The checkpointLocation is fine as configured, because the DBFS root of the workspace is the recommended default location for production stream state","The checkpointLocation is irrelevant, because checkpoints are optional when the stream writes its output into a Delta table","The checkpointLocation can be any random path, because Auto Loader rebuilds an equivalent checkpoint from the target table on every run"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/best-practices","referenceTitle":"Auto Loader best practices — Databricks Docs"}}
{"id":"RAN-DBDEA-00125","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:eb32f9c29c55eaf2dcf41c573ea1ed89432b01675cad18362f970e93036b52c9","aliases":[],"snapshot":{"statement":"A team's ingestion notebooks currently reference raw files by full s3:// URIs, with long-lived cloud credentials pasted directly into the code. The governance program now requires that every landing location be a centrally governed path, that no cloud credentials appear in notebooks, and that access be granted and revoked through one central permission model. Where should the landing files be stored?","options":["In the current bucket as is, with instance profiles referenced inside each notebook and updated by hand whenever the cloud credentials rotate","In the bucket through DBFS mounts configured with cloud credentials on the cluster, the traditional pattern for production landing areas","In a Unity Catalog volume, using /Volumes/catalog/schema/volume paths governed by READ VOLUME and WRITE VOLUME privileges","In a public bucket with anonymous access enabled, so the notebooks can keep using plain s3:// URLs without credential management"],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00126","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7b3f10e7a974805c63e79ec496472ce14e546cca1181ada28569c476c98e1577","aliases":[],"snapshot":{"statement":"One shared landing directory receives deliveries from many producers: clickstream exports, audit logs, invoice images, and order events, all mixed in the same path. Only the files named with the pattern orders-*.csv may ever reach the orders table; everything else must be ignored automatically, with no manual file handling between runs. Which ingestion approach meets this requirement?","options":["A COPY INTO statement with a PATTERN clause of orders-*.csv, so only the files whose names match the pattern are loaded","A staging load of every file in the directory, followed by a DELETE that removes the rows originating from non-order files","A manual archive step that moves every non-order file to another prefix before each run, leaving only the matching files in place","An Auto Loader stream with cloudFiles.pathGlobFilter set to orders-*.csv, so only files matching the glob pattern are ingested"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00127","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:92cf0f3f097461da4c8d34ca0569f1ae415a0d0dbd5cd865b18e04708cce66f9","aliases":[],"snapshot":{"statement":"A Lakeflow Job task calls a partner REST API that uses OAuth 2.0 client-credentials authentication: the task presents a client ID and a long-lived client secret to obtain a short-lived access token at runtime. The security team requires that the secret be stored in one managed place, referenced from the task without the value appearing in run history or logs. Where should the client secret be stored?","options":["As a string parameter on the Lakeflow Job, so the task reads the secret from its parameter list at the start of every run","In a Databricks secret scope, fetched at runtime with dbutils.secrets.get, which redacts the value in output and logs","In an environment variable set in the cluster's Spark configuration, where it is readable by anyone who can attach to that cluster","In the Git folder repository that holds the pipeline code, with branch permissions restricted to the data engineering team"],"answer":"B","reference":"https://docs.databricks.com/aws/en/security/secrets","referenceTitle":"Manage secrets — Databricks Docs"}}
{"id":"RAN-DBDEA-00128","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9f5fb8af934b1f261e0afd39582f28e95acc49ae639184fd8d157e3fc7ca0bd8","aliases":[],"snapshot":{"statement":"An operations dashboard queries a Delta table that must never be more than a few minutes out of date. The team runs an Auto Loader stream on always-on compute and wants a new micro-batch to run every 60 seconds, around the clock, without manual intervention. Which trigger configuration produces that behavior?","options":["Use .trigger(processingTime='1 minute'): the query schedules continuous micro-batches, each starting about one minute after the previous","Use .trigger(availableNow=True) with a one-minute interval: the stream processes whatever is available and then starts again every minute","Use .trigger(processingTime='1 minute') for discovery only: Auto Loader scans for files faster, while micro-batches follow the default cadence","Use no trigger at all: streams default to batch execution, so the table only advances when someone runs a manual refresh command"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00129","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:8cda6b2b858565b027a58bd35ad56accd440c0cc3bdaf40e3dc7622e48d4e53c","aliases":[],"snapshot":{"statement":"A company runs several Lakeflow Connect managed connectors, and three of them read from PostgreSQL databases that all accept the same service-account credential. The platform team wants to store that credential exactly once, reference it from each connector, and guarantee that engineers who configure or use the connectors can never see the secret value itself. Where do Lakeflow Connect managed connectors hold their source credentials?","options":["In a workspace-backed secret scope that every engineer who configures a connector must be granted permission to read","In plaintext variables defined in the team's Automation Bundle and versioned in Git together with the pipeline definitions","In a Unity Catalog connection created with CREATE CONNECTION; connectors reference it and the stored secret stays hidden","In a personal access token issued to each engineer, shared with teammates whenever someone joins or leaves the team"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/lakeflow-connect/","referenceTitle":"Lakeflow Connect connector concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00130","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:643cc03c77acc2ef2c2b8d8b11a15ef9c76d9ae3f00a9fe5cb4d4eda6cd568b1","aliases":[],"snapshot":{"statement":"A streaming table created with read_files over a landing directory has been live for weeks. A faulty transformation in the pipeline polluted its contents; the source files are still intact in cloud storage, and the transformation has now been fixed. The team decides to rebuild the table with a full refresh. Which statement describes what REFRESH TABLE ... FULL REFRESH does?","options":["It re-executes only the micro-batches whose output the faulty transformation affected, leaving every correctly loaded batch untouched in the table","It discards the streaming state that tracks processed data and reprocesses everything from the source, fully re-deriving the table","It drops and recreates the table schema before reloading, resetting column types and constraints to their original definitions","It fires automatically whenever the source schema changes, so issuing it manually after fixing a transformation is redundant"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ldp/concepts/streaming-tables","referenceTitle":"Streaming tables (Lakeflow pipelines) — Databricks Docs"}}
{"id":"RAN-DBDEA-00131","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:1740502a8de2524d29af711699baf8607caf29d53be36178e29f44be8c8b76fc","aliases":[],"snapshot":{"statement":"A scheduled Lakeflow Job runs this statement every night to load the day's files into a reporting table:\n\n```sql\nCOPY INTO main.ops.deliveries_raw\nFROM '/Volumes/main/ops/landing/deliveries/'\nFILEFORMAT = CSV;\n```\n\nLast night the run failed halfway through, when it hit a corrupted file after several valid files had already been read. What is the state of the target table this morning?","options":["Unchanged: COPY INTO is atomic, no partial data was committed, and a rerun after the fix loads all still-pending files","Partially loaded: the valid files read before the corruption were committed, and a rerun resumes from the file after the corrupted one","Empty: the failed statement rolled the table back to its creation state, and the next successful run reloads everything from scratch","Locked: the failed run marked every listed file as loaded, so the team must purge the copy history before it can retry"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00132","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:873a62ca902649fb6d86bde613bb38cb56e19d88d29a8cfb3aab409dde058461","aliases":[],"snapshot":{"statement":"A company's gold layer ends with a nightly aggregated orders table that a legacy planning system can only consume from its own PostgreSQL database. Right after the aggregation finishes each night, the table's full contents must be delivered into a PostgreSQL table that the legacy system reads. How should the team deliver the gold table into PostgreSQL?","options":["Run COPY INTO against the PostgreSQL table, because COPY INTO moves data between Delta tables and any external SQL database","Share the gold table using Delta Sharing: the open protocol automatically materializes it as a table inside the PostgreSQL database","Schedule a notebook that displays the aggregation results and exports them to CSV, letting the legacy team load the file by hand","Add a task that writes the DataFrame with .format('jdbc'), passing the PostgreSQL URL, credentials, and save mode as options"],"answer":"D","reference":"https://docs.databricks.com/aws/en/archive/connectors/jdbc","referenceTitle":"Query databases using JDBC — Databricks Docs"}}
{"id":"RAN-DBDEA-00133","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:19850b597c981cad805b111951170b0cb73914b42d4583cbc399a66b8287c958","aliases":[],"snapshot":{"statement":"During a performance review of a silver pipeline, a reviewer notices that the author registered a Python UDF to uppercase a country code and to split a full name, although both transformations are available as built-in Spark functions. The author asks why the reviewer insists on replacing the UDF with the native functions.\n\nWhich explanation is correct?","options":["Built-in functions execute in the optimized engine with no Python serialization and remain visible to the optimizer.","Python UDFs are always faster than built-in functions because they run dedicated code compiled for each transformation.","Built-in functions are exposed only in SQL; PySpark DataFrames must go through a Python UDF for equivalent logic.","Performance is identical for both approaches, so replacing the UDF is purely a matter of coding style preference."],"answer":"A","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00134","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:0d51c0f3b5a7462be5f5204343e5d150327ee6c83e81f2885a7e2de67175cc99","aliases":[],"snapshot":{"statement":"A PySpark job reads a Delta table into a DataFrame that currently has 200 partitions spread across the cluster. Before writing the result, the team wants to end up with 8 partitions to avoid producing hundreds of tiny output files, but without paying for a full shuffle of the data.\n\nWhich operation should they use?","options":["repartition(8), the only DataFrame operation able to reduce the number of partitions","coalesce(8), which always preserves the current data distribution across the worker nodes","coalesce(8), which merges existing partitions into fewer ones without a full shuffle","either operation, because repartition and coalesce are synonyms that differ only in naming"],"answer":"C","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00135","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:bda5fd7d7672c11d5a24b00d1b4b1bd25c7ed690da7d0395372be9df720e61ac","aliases":[],"snapshot":{"statement":"A junior engineer chains several DataFrame operations in a notebook: two withColumn calls, a filter on status, and a select of five columns. Each cell returns almost instantly, and she is surprised that the source table is only read minutes later, when she finally calls count(). She asks her lead what happened.\n\nWhich explanation is correct?","options":["Each transformation executed immediately and stored its result; count() only aggregated the precomputed data.","Transformations are lazy and only extend the logical plan; count() is an action that triggers execution.","Actions are lazy as well; only show() forces Spark to execute the plan and read the source data.","The accumulated plan is executed automatically at cluster shutdown, when pending work is flushed to storage."],"answer":"B","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00136","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:03401f9e3d9828f0012259bad0843e9300bc8c365e65f66066dc343bb19bbda6","aliases":[],"snapshot":{"statement":"Two analysts on different teams must share the same SQL definition — a curated customer list built over several silver tables. The definition must live in one governed place: secured with Unity Catalog permissions, centrally versioned, and readable from any workspace attached to the metastore. The analysts use separate notebooks on separate compute.\n\nWhich object meets these requirements?","options":["A temporary view (CREATE TEMP VIEW), which lives only inside the session that defines it","A global temporary view (CREATE GLOBAL TEMP VIEW), shared across sessions on the same cluster","A .sql script in a shared folder, which each analyst must open and run in her own session","A view created with CREATE VIEW in a Unity Catalog schema, stored in the metastore and governed with grants"],"answer":"D","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-syntax-ddl-create-view","referenceTitle":"CREATE VIEW — Databricks Docs"}}
{"id":"RAN-DBDEA-00137","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:26aab568cb2c5cba83194f660c0506b531a167c4ed8c54722004eac9c4c6f957","aliases":[],"snapshot":{"statement":"Last night, a batch job ran a faulty UPDATE on catalog.sales.orders, overwriting the status column on every row. The faulty commit landed as version 3, and the table is now at version 4. An analyst must produce a report from the table exactly as it existed before that commit, without changing the current data.\n\nHow should the analyst read the pre-UPDATE snapshot?","options":["Query the historical snapshot with SELECT * FROM catalog.sales.orders VERSION AS OF 2","Run RESTORE TABLE catalog.sales.orders TO VERSION AS OF 2, then execute the report again","Run the report on the current table with a WHERE clause that filters out the overwritten rows","Read the table's underlying Parquet data files directly from the storage location with Spark"],"answer":"A","reference":"https://docs.databricks.com/aws/en/delta/","referenceTitle":"Delta Lake guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00138","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:41f4e6411d2eee6c15aee795c9789864e94adb003a7efb17e3d248f89dde9194","aliases":[],"snapshot":{"statement":"An auditor must verify catalog.finance.ledger as it stood 20 days ago. The team runs the time travel query for that snapshot, but it fails with a FILE_NOT_FOUND exception pointing at a data file. The table is vacuumed weekly with the default retention settings, and OPTIMIZE also runs regularly.\n\nWhat is the most likely cause of the failure?","options":["Time travel preserves every version since table creation, so the failure must come from missing SELECT privileges","VACUUM only compacts small files into larger ones and never deletes any file from storage","VACUUM removed data files older than the default 7-day retention, making older snapshots unreadable","RESTORE TABLE can recover any prior version even after VACUUM has removed the underlying files"],"answer":"C","reference":"https://docs.databricks.com/aws/en/delta/best-practices","referenceTitle":"Best practices: Delta Lake — Databricks Docs"}}
{"id":"RAN-DBDEA-00139","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f03d618a96b7019de3f8274fef63998f730e48e1933636d7d098b303390715a0","aliases":[],"snapshot":{"statement":"A streaming pipeline commits new data to catalog.ops.events every 30 seconds. After a few months, the Delta table holds hundreds of thousands of small files, and dashboard queries have degraded: execution time is dominated by listing and opening files rather than scanning data.\n\nWhich maintenance action restores read performance?","options":["Run VACUUM with the default retention settings, so that small files are merged into larger ones","Run OPTIMIZE, which compacts the small files into larger ones for more efficient reads","Rewrite the entire table every night with repartition(1) to produce one giant file","Scale up the compute behind the dashboards while leaving the file layout untouched"],"answer":"B","reference":"https://docs.databricks.com/aws/en/delta/best-practices","referenceTitle":"Best practices: Delta Lake — Databricks Docs"}}
{"id":"RAN-DBDEA-00140","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ed10e4158bfa78a04938b9e78d97b11b29d94310b8333460c061574becab85e8","aliases":[],"snapshot":{"statement":"Every dashboard query against catalog.sales.orders derives the order date by calling to_date(order_ts) — an expression repeated across dozens of reports over billions of rows. The team wants the derived value computed once, at write time, and physically stored as a column of the table.\n\nWhich Delta Lake feature provides this?","options":["A view that computes to_date(order_ts) on every read of the underlying table","A trigger that fires AFTER INSERT on the table and updates the derived column","A dashboard-level expression, so that the calculation stays out of the storage layer","A generated column: order_date DATE GENERATED ALWAYS AS (CAST(order_ts AS DATE))"],"answer":"D","reference":"https://docs.databricks.com/aws/en/delta/","referenceTitle":"Delta Lake guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00141","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:97ac479bebdd2af8aab9cbad916c2b231adb3921f671ba362d2a68ef377fea3a","aliases":[],"snapshot":{"statement":"An operational database streams its change feed into a bronze table: each row carries the operation (insert, update, delete), a monotonically increasing sequence number, and the changed columns. In a Lakeflow Declarative Pipeline, the team must keep a streaming table in sync with the current source state — deletes included — without writing manual merge logic.\n\nWhich construct should the pipeline declare?","options":["A hand-written MERGE INTO statement that joins the change rows to the target on the primary key","An AUTO CDC INTO flow with KEYS and SEQUENCE BY, replacing the former APPLY CHANGES syntax","A row-level expectation that drops invalid change rows as they flow into the target table","A full refresh that recreates the streaming table from scratch whenever a new batch arrives"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ldp/cdc","referenceTitle":"The AUTO CDC APIs (Lakeflow Declarative Pipelines) — Databricks Docs"}}
{"id":"RAN-DBDEA-00142","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6444d7671e6640249b0825cabe3097feae9b84fe873984f62eac92c0b4326959","aliases":[],"snapshot":{"statement":"A Delta table catalog.sales.payments declares amount as DOUBLE. A new ingestion job builds a DataFrame in which amount is still STRING, with raw values like '125.50', and appends the DataFrame to the table. The job author asks what Delta Lake will do when the write is attempted.\n\nWhat is the expected behavior?","options":["The write fails: Delta enforces the table schema and rejects a column type that does not match","The write succeeds, storing STRING values that are converted back to DOUBLE whenever readers query","Delta automatically adds a second column, amount_string, to hold the mismatched incoming values","Delta writes only the columns whose types are compatible and silently drops the mismatched one"],"answer":"A","reference":"https://docs.databricks.com/aws/en/delta/","referenceTitle":"Delta Lake guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00143","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:e025f3d07b541243bc1a5e3c8c943dd709084f49658aa19f3581468b23e4530b","aliases":[],"snapshot":{"statement":"Before promoting a new PySpark transformation into the nightly production job, an engineer wants to see what Spark intends to do with it — the planned join strategy, the filters, and the projection — without spending cluster resources actually running the query.\n\nWhich action gives her this preview?","options":["Call count() after every transformation and compare the elapsed time of each stage","Nothing — lazy evaluation makes the plan unknowable until the query has fully executed","Call df.explain() to print the logical and physical plans without executing anything","Run the job in production and inspect the Spark UI, the only place where plans appear"],"answer":"C","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00144","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:3888c9c788a3da036829568a5fa104102ea23b0aa7992380484bc5101e00a922","aliases":[],"snapshot":{"statement":"A reconciliation job must align catalog.sales.silver_orders with the source extract catalog.sales.latest_orders in one atomic pass: matched orders get their status refreshed, matched orders whose source status is 'CANCELLED' must be removed from the target, and unmatched source orders must be inserted.\n\n```sql\nMERGE INTO catalog.sales.silver_orders AS t\nUSING catalog.sales.latest_orders AS s\nON t.order_id = s.order_id\n-- WHEN clauses omitted\n```\n\nWhich approach meets all three requirements in this single atomic pass?","options":["Run two MERGE statements in sequence: one that updates matched rows and inserts new ones, then one that deletes the cancelled orders","Use a single MERGE with WHEN MATCHED THEN UPDATE SET * and WHEN NOT MATCHED THEN INSERT *, issuing a separate DELETE for the cancelled orders","Use a single MERGE with WHEN MATCHED AND s.status = 'CANCELLED' THEN DELETE, then WHEN MATCHED THEN UPDATE SET *, then WHEN NOT MATCHED THEN INSERT *","Replace the MERGE with an UPDATE on the matched rows followed by REPLACE WHERE, rewriting the table with the surviving and new orders"],"answer":"C","reference":"https://docs.databricks.com/aws/en/delta/merge","referenceTitle":"Merge data (Delta Lake upsert) — Databricks Docs"}}
{"id":"RAN-DBDEA-00145","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:665430832c46ba4d44a2d23fb687e138719a2f4520c24c39e4b98809fd439975","aliases":[],"snapshot":{"statement":"Over the past week, catalog.ops.shipments has behaved strangely: row counts shift overnight even though the team's own jobs were idle, and the table version keeps climbing. The engineer needs to determine exactly which operations were committed to the table, when, and by which user or service principal.\n\nWhich command gives her this commit-level audit trail?","options":["SHOW TABLES with a filter on the last modified date, which lists only recent writes","A manual read of the _delta_log JSON transaction log files parsed directly from storage","The workspace audit log, which is described as the only existing record of table writes","DESCRIBE HISTORY, which lists every commit with its operation, timestamp, and user"],"answer":"D","reference":"https://docs.databricks.com/aws/en/delta/","referenceTitle":"Delta Lake guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00146","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a15fde795c7843924c082f6b9cc9bc85d3e26a3ca8e64e49d878e1404c993118","aliases":[],"snapshot":{"statement":"The source system exports two timestamps as STRING columns: ordered_at and shipped_at, both formatted like '2026-07-02 14:45:00'. In the silver transformation, the team must add fulfillment_days with the number of calendar days between the order date and the ship date, written in Spark SQL.\n\nWhich approach should they use?","options":["Parse both columns with to_timestamp(col, 'yyyy-MM-dd HH:mm:ss'), then compute datediff(shipped_at, ordered_at)","Slice fixed character positions with substring() to build year, month, and day, then subtract the dates","Convert both columns with unix_timestamp(), subtract, and divide the difference by 86400","Register a Python UDF, because date parsing requires a library that Spark SQL lacks natively"],"answer":"A","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00147","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6ce577972d9dee92fb8164d670520e3c4cfbbe85dea022af9d4a1c5eac516bf0","aliases":[],"snapshot":{"statement":"The table catalog.sales.order_items stores one row per line item, with columns order_id and item. An export job must return exactly one row per order, carrying all of its items in a single ARRAY<STRING> column named items — a native array, not a concatenated string.\n\nWhich query shape produces this result?","options":["Explode the item column first, then concatenate the values into one delimited string","SELECT order_id, collect_list(item) AS items FROM catalog.sales.order_items GROUP BY order_id","Pivot the rows into one explicit column per item position, such as item_1 and item_2","It is impossible in pure SQL, because aggregate functions cannot return complex types like arrays"],"answer":"B","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00148","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:bc8a039886c873819891b97fa92db7763226154a5dd10492a35afb55cf6cf403","aliases":[],"snapshot":{"statement":"A production Lakeflow Job that publishes Unity Catalog tables was created by a senior data engineer who left the company last month. His user account has since been removed, and the job's nightly runs now fail with permission errors when reading the source tables. The job's code and schedule are correct.\n\nWhat should the team change to make the job independent of any person's account?","options":["Set the job's Run as identity to a production service principal that holds the required table grants","Set Run as to the new team lead now, and to whoever leads the team after each reorganization","Leave Run as empty, since production jobs automatically run with the permissions of whoever triggered them","Grant the departed engineer's user account the workspace admin role so that its permissions are never removed again"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/privileges","referenceTitle":"Manage permissions on Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00149","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:1d1beb2a6dd28744871180fe8b5d474aec03c62443d4e010ffe4c9fef3e4f567","aliases":[],"snapshot":{"statement":"A team maintains a shared Python module, data_prep.py, which lives in the team's Git repository and is reviewed through pull requests. Several production pipelines depend on it. They are adding a task to a Lakeflow Job that must execute this file exactly as it exists on the main branch, with no manual copies.\n\nHow should the task be configured?","options":["Paste the module's current code into a notebook task and refresh that copy by hand whenever the module changes","Copy data_prep.py into the workspace first, because Lakeflow Jobs can only execute notebook tasks","Use a Python script task with the Git provider as its source, pointing to data_prep.py on the main branch","Run the script manually in a terminal on an attached cluster and load the exported CSV files afterward"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/configure-task","referenceTitle":"Configure and edit tasks in Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00150","exam":"DBDEA","topics":["jobs"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:87901b43d94d61aa20fd3130eee973b472daca96c60806b17549fd545c3bdddc","aliases":[],"snapshot":{"statement":"A data engineering team must run a Lakeflow Job at 02:00 every weekday, Monday through Friday, so that reporting tables are fresh before analysts arrive. In the job's schedule settings, they open the cron expression editor to type the schedule by hand.\n\nWhich expression should they enter?","options":["0 2 * * MON-FRI","0 0 2 * * 1-5","MON-FRI 02:00","0 0 2 ? * MON-FRI *"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00151","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:a685e7b3c901b5c30862830123c1006ac157c796234038808cab0f4422dac9df","aliases":[],"snapshot":{"statement":"A task in a production Lakeflow Job got stuck in an infinite loop over the weekend. Nobody was watching, and the job stayed in the running state for three days, holding compute and blocking downstream pipelines. The team wants a stuck run like this to be stopped and marked as failed automatically.\n\nWhat should they configure?","options":["Nothing — Lakeflow Jobs offer no timeout setting, so a stuck run must be detected and canceled manually","The job cluster's auto-termination setting, which terminates compute that has been idle for the configured minutes","A timeout on scheduled runs only, because runs started manually with Run now are allowed to run indefinitely","A Timeout setting on the job or the task, so that exceeding the limit stops the run with a Timed Out status"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00152","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c62cd73e5cf638bb17fcb7845356a4f68017f8dbf77595115f87566037abb38d","aliases":[],"snapshot":{"statement":"When a nightly Lakeflow Job fails, the team already receives an email notification. Operations, however, also needs every failure recorded in its own incident system, which exposes an HTTP endpoint, and the on-call rotation must be paged through PagerDuty. All of this should happen automatically as part of the run.\n\nHow can the team meet this requirement?","options":["Keep email as the notification, since Databricks cannot target external systems, and let the incident system poll the jobs API","Add a webhook destination to the job's failure notification, so that Databricks posts a JSON payload to the endpoint","Add a final task that calls the incident system's endpoint itself at the end of every run","Switch the job to a file arrival trigger, because webhook notifications only work for event-triggered jobs"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/notifications","referenceTitle":"Add notifications on a job — Databricks Docs"}}
{"id":"RAN-DBDEA-00153","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9f94049cc74429132013273f1ea488848981438a931cb6335febe80c54e53d1d","aliases":[],"snapshot":{"statement":"The production version of an ETL notebook lives on the release branch of the team's GitHub repository. Today, a team member copies the notebook into the workspace by hand before each release, and last week's job ran a stale copy that missed two reviewed fixes. The next release must run exactly the reviewed code, with no manual steps.\n\nHow should the notebook task be configured?","options":["Keep the manual copy, but add a release checklist that verifies the workspace version before each run","Leave the task pointing at the workspace, because jobs can only execute notebooks stored there","Set the task's Source to the Git provider and point it at the notebook on the release branch","Point the task at a /Repos workspace path and trust each developer to pull the release branch before the run"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/tasks/notebook","referenceTitle":"Notebook task for jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00154","exam":"DBDEA","topics":["jobs"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:21c15df7730b2c5d9503667d1946b2eea08bce279b5f0b7c3a579bcfcc91621a","aliases":[],"snapshot":{"statement":"A Lakeflow Job runs ingest, transform, and publish as a linear chain. The team must add an alerting task that runs only when the ingest task fails — it must never run when the pipeline succeeds. The alerting code is ready; only the task's wiring is missing.\n\nHow should the alerting task be wired into the job?","options":["Depends on ingest with the Run if condition All failed, so it runs only when ingest has failed","No dependencies — it runs in every run, queries the previous run's outcome through the API, and exits early on success","Depends on publish with the default Run if condition, closing the chain after every successful run","A second, identical job that contains only the alerting task and is triggered by hand when the first job fails"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/control-flow","referenceTitle":"Control the flow of tasks within Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00155","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:64e174499ea3c696386b5ed17863e94d77a8516eaf689f971053202157e1b510","aliases":[],"snapshot":{"statement":"A daily Lakeflow Job must fire at 06:00 in the company's Sao Paulo office (America/Sao_Paulo). The platform team in Paris created the schedule, so its timezone is Europe/Paris and its cron fires at 11:00 — 06:00 in Sao Paulo during the European summer. In late October, Europe ended daylight saving time, and the job began firing at 07:00 local.\n\nHow does the team make the firing time deterministic?","options":["Set the schedule's timezone to America/Sao_Paulo, so the cron is always resolved in Sao Paulo time","Keep Europe/Paris and edit the cron hour by hand twice a year, when European clocks change","Set spark.sql.session.timeZone to America/Sao_Paulo in the job's tasks, so the scheduler fires in local time","Leave the timezone alone, because a job's cron automatically follows the local timezone of the engineer who created it"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00156","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e5d1e81a6ae00682bb7acae2391fe11da6d1996bfbc4da76d1c07a325e71aa69","aliases":[],"snapshot":{"statement":"A Lakeflow Job orchestrates an end-to-end flow: a notebook task lands raw files, then a Lakeflow Declarative Pipeline must process the new data, and a final task exports the curated tables. The team is now adding the middle step to the job.\n\nHow does the pipeline take part in this job?","options":["As a Pipeline task, which always triggers a full refresh of the pipeline before processing","It cannot, because pipelines run only as standalone objects with their own schedules","As a notebook task that re-runs the pipeline's source notebook inside the job","As a Pipeline task that selects the pipeline and triggers a refresh — incremental by default"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/configure-task","referenceTitle":"Configure and edit tasks in Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00157","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ee11d4746fb9b8e26714f4afc67972dca313d3d6fa6910afc231311d4d6598dc","aliases":[],"snapshot":{"statement":"A nightly Lakeflow Job takes a processing_date parameter whose default is the current date. Late this evening, an on-call engineer must reprocess yesterday's data exactly once, and tomorrow's scheduled run must continue using the default. The job definition itself may not change.\n\nWhat should the engineer do?","options":["Edit the job's default parameter to yesterday's date, start the run, and revert the default before the nightly schedule fires","Clone the job into a one-off reprocess version with the new date, run it, and delete the clone afterwards","Start the run with Run now with different parameters, passing processing_date for that run only","Deploy a new bundle target that sets the parameter, since parameter values can only change through a deployment"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00158","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:82b72199c0999582a6c5df5b56b8a9d0fe274859619bb3cc0100b5264671e5e9","aliases":[],"snapshot":{"statement":"A Lakeflow Job's transformations require pandas 2.x. The team has been prototyping on a shared cluster that other teams also depend on, and that cluster has pandas 1.5 pinned — upgrading it would break the other teams' workloads. The job is about to go to production.\n\nWhere should pandas 2.x be installed for this job?","options":["On the shared cluster, after warning the other teams and having them retest their pipelines","With %pip install at the top of the job's notebook, which fixes the version for every job the same way","Nowhere, because pinned libraries are not supported on job clusters, so the code must avoid pandas 2.x","On the job's own job cluster, whose library list is isolated from the compute other teams share"],"answer":"D","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00159","exam":"DBDEA","topics":["cicd"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:1a0085c439409a547af7ed42a6c21182837ca574410a0011ebc5acef5ce6e9d5","aliases":[],"snapshot":{"statement":"A newly hired data engineer wants to start adopting Automation Bundles for the team's Lakeflow Jobs and pipelines. They need a first bundle project on their laptop — a databricks.yml plus starter resource definitions — but have never written bundle YAML and do not want to assemble the project files by hand. Which action gets them a working scaffold the fastest?","options":["Copy a full databricks.yml example out of the Databricks documentation and adapt its fields by hand each time the team starts a new project","Ask a workspace administrator to create the bundle in the workspace UI, because bundles cannot be created locally","Connect a remote Git repository to the workspace first, because databricks bundle init will not run without an existing repository","Run databricks bundle init, pick one of the Databricks-provided default templates, and answer the prompts that generate the project files"],"answer":"D","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00160","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:745233bdda8a2bfa315e8b00a57390b47e3b5a7b1390d4d4fb4adc2bd9844abd","aliases":[],"snapshot":{"statement":"A team's CI pipeline deploys every open pull request as its own Automation Bundle target — pr_142, pr_143, and so on — creating a disposable copy of the pipeline's Lakeflow Job and notebooks in a shared dev workspace. Pull request 142 has just been merged, and the pipeline must clean up that PR's deployment completely, leaving no orphaned jobs, notebooks, or artifacts, without touching other pull requests' resources. What should the merge step run?","options":["Delete the shared dev workspace after every merge and let the next pull request deployment recreate everything inside it","Run nothing in the merge step — the PR deployment's job cluster auto-terminates after its runs, removing the deployed resources with it","Run databricks bundle destroy and accept its side effect of permanently deleting the bundle's source files and Git repository","Run databricks bundle destroy -t pr_142, which deletes the job, files, and artifacts that this target's deployment created"],"answer":"D","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00161","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f42f334afe7b0db12ce7ab15591629650d5801bd10a435c1b1c34ff48d7b6a2d","aliases":[],"snapshot":{"statement":"A team deploys one Automation Bundle to dev and prod targets. In dev only, the group data_team must hold CAN_MANAGE_RUN on the bundle's Lakeflow Job so members can trigger and monitor runs; in prod the group must receive nothing. The team wants the grant applied automatically on every deploy and reviewed in Git like the rest of the configuration. How should this be set up?","options":["Add a permissions entry granting data_team the CAN_MANAGE_RUN level on the job resource within the dev target of the bundle","Have an administrator grant CAN_MANAGE_RUN to data_team from the job's Permissions tab in the dev workspace UI after every deploy","Make data_team a workspace admin in the dev workspace so its members can run every job there without bundle changes","Skip permissions in the bundle, because bundle permissions mappings can only grant view-level access such as CAN_VIEW"],"answer":"A","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00162","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0ae72a7af7bff6c4d371604f5389a37806f5a3faf508f6abaf18d63de85d0c27","aliases":[],"snapshot":{"statement":"A team's Automation Bundle has grown into a single 600-line databricks.yml: two Lakeflow Jobs, one pipeline, and per-environment overrides for dev, test, and prod all live in one file that reviewers struggle to read. The engineers want resource definitions in their own files, with each target spelling out only what differs from the root definitions. How should the bundle be restructured?","options":["Keep everything in one databricks.yml, because the bundle specification recommends keeping all resource definitions and targets together in a single configuration file","Move the job and pipeline definitions into resource files listed in the include mapping; targets inherit root definitions and override only what changes","Create a separate bundle per environment, duplicating the job and pipeline definitions in each one to keep the files short","Give up on splitting the file: a bundle cannot reference other files, so the single databricks.yml must simply be reorganized internally"],"answer":"B","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00163","exam":"DBDEA","topics":["cicd"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:a7cdc2e702fa7597314ccff0d1e9fe2ee6aaf8620d8485509fe6674186806576","aliases":[],"snapshot":{"statement":"A nightly GitHub Actions workflow authenticates as the service principal ci-deployer and runs databricks bundle deploy -t prod, creating and updating the team's Lakeflow Jobs and pipelines in the production workspace. Security policy requires machine identities to hold the narrowest access that still lets the pipeline deploy successfully. Which privilege assignment satisfies the policy?","options":["Make ci-deployer an account admin, so the nightly deploy can never fail because of a missing permission","Rely on inheritance: a service principal automatically holds the permissions of the engineer who authored the bundle","Grant ci-deployer only what the deployment needs: privileges to create and manage its jobs and pipelines, plus required catalog and schema access","Assign no privileges at all, because bundle deployments run under the workspace identity rather than the signed-in principal"],"answer":"C","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00164","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7a15e481f69c83961f250fe7577371874d5e95c6879e5641c34f6ddc248669b7","aliases":[],"snapshot":{"statement":"An analytics team serves executive dashboards from a Medium SQL warehouse. During the morning peak, individual dashboard queries take over a minute, and the query profile shows that each query itself is resource-hungry, with high CPU and memory use per node; only a handful of queries run concurrently. The team wants each query to run faster. What should they change on the warehouse?","options":["Create one dedicated warehouse per dashboard and manually route refreshes to whichever is idle","Increase the warehouse size so every query runs with more memory and CPU per node","Lengthen the auto-stop delay so the warehouse stays warm between dashboard refreshes","Recreate the warehouse from scratch, because its size is fixed at creation and cannot be edited"],"answer":"B","reference":"https://docs.databricks.com/aws/en/compute/sql-warehouse/warehouse-behavior","referenceTitle":"SQL warehouse sizing, scaling, and queuing behavior — Databricks Docs"}}
{"id":"RAN-DBDEA-00165","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:36f90b61a16a9c3fddfa9214cf841becb48600d4befdcd8119fc331ce654a646","aliases":[],"snapshot":{"statement":"A scheduled notebook task fails with an uncaught Python exception. The job run page shows only the exception type and a truncated message, and the engineer needs the complete traceback together with the driver's stdout and stderr output to debug it. Where should the engineer look?","options":["In the cluster's driver logs, opened from the Driver Logs tab or from delivered log storage","Only on the engineer's local machine, because driver stdout and stderr are never retained by the platform","In the Spark UI's Jobs and Stages tabs, which capture the complete Python stack trace of the failure","In the cluster metrics charts, because Ganglia-style dashboards record application-level tracebacks"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00166","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:0169116747f1610fc41bfe411c490869a214fdd00b62f4c96c9eb4c7b5d9d366","aliases":[],"snapshot":{"statement":"A Databricks SQL query that joins two gold tables and aggregates them takes minutes every night. The engineer wants a visual diagnosis: which operators dominate the elapsed time, how many rows each step produces, and whether a full table scan or an exploding join is to blame. How can the engineer get this analysis?","options":["Open the Lakeflow pipeline event log, because query profiles are generated only for pipeline queries","Enable verbose mode on the warehouse first; without it the query profile never displays any operators","Open the query profile from query history: it shows the operator DAG with time, rows, and memory","Re-run the query so it is served from the query cache, then open the profile of that cached execution"],"answer":"C","reference":"https://docs.databricks.com/aws/en/sql/user/queries/query-profile","referenceTitle":"Query profile — Databricks Docs"}}
{"id":"RAN-DBDEA-00167","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:63188e4a7d49df704a25fc86fd3f20cc678f51d2a4381f984e160b5c7318328b","aliases":[],"snapshot":{"statement":"A company's nightly Lakeflow Job cluster failed to launch: the cluster event log shows a cloud provider error stating that no instances of the requested type are available in the selected availability zone. The same instance type launched fine last week, and the job must run reliably every night. How should the team mitigate this failure?","options":["Reinstall the cluster's libraries from scratch, because a corrupted dependency can block node allocation","Raise the auto-termination threshold so idle nodes remain available for the next night's launch","Audit the init script's permissions, since capacity errors are always caused by failed init scripts","Allow a mix of spot and on-demand instances, more availability zones, or alternate instance types"],"answer":"D","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00168","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:daa11cd1bb8e6fc9a46bb57c1d5a14a4b92a5ef3021414d23d94f23c1386fb2a","aliases":[],"snapshot":{"statement":"A data engineering team runs a nightly workload of SQL queries and Spark DataFrame transformations over large Delta tables — heavy scans, joins, and aggregations. Their cluster uses a Databricks Runtime with Photon enabled, and a new engineer asks what Photon actually contributes. Which answer is correct?","options":["A vectorized C++ engine that accelerates SQL and DataFrame workloads such as scans, joins, and aggregations","A compiler that vectorizes arbitrary Python UDFs so they run as fast as built-in expressions","A columnar file format that competes with Delta Lake for the on-disk layout of tables","A streaming-only accelerator that applies exclusively to Auto Loader and streaming tables"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute/photon","referenceTitle":"Photon — Databricks Docs"}}
{"id":"RAN-DBDEA-00169","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:be696ca04da9883899ff0b3d1b11cdf1e660aa98169a4f91fa999a71d637eeb6","aliases":[],"snapshot":{"statement":"A Lakeflow Job runs on an autoscaling job cluster configured with minimum 2 workers and maximum 8. At the start of a large backfill, thousands of tasks are queued, but the cluster runs with only 2 workers for several minutes before more nodes join. The engineer expected 8 workers immediately. What explains the behavior?","options":["Autoscaling always jumps straight to the configured maximum worker count as soon as the job run starts","Autoscaling adds nodes gradually as task queues and utilization justify them; the minimum is only a guaranteed baseline","The maximum workers setting is ignored on job clusters, so this cluster can never exceed two nodes","Autoscaling does nothing by itself: an engineer must call the Clusters API and attach nodes manually as the backlog grows"],"answer":"B","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00170","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f935ce6e91535cda0b54294d9c03369beb304ab6d51395d9f6819fb710951119","aliases":[],"snapshot":{"statement":"A data engineer must register an existing Delta dataset as an external table in Unity Catalog and runs the following statement:\n\n```sql\nCREATE TABLE main.sales.orders_ext (order_id BIGINT, amount DECIMAL(10,2))\nUSING DELTA\nLOCATION 's3://acme-raw/orders/';\n```\n\nThe statement fails with an error saying that the input path does not match any external location. What is the most likely cause?","options":["The s3:// path would be accepted if the bucket were publicly readable, so the failure means the bucket policy must be opened.","The engineer's account is missing SELECT on the main catalog, and granting SELECT on a catalog is what authorizes table creation.","The path is not covered by any registered external location, an object that references a storage credential.","The statement was run on a shared cluster, and s3:// locations in CREATE TABLE only resolve on single-user clusters."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00171","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:64a4b4b13bab69c6baa411cc26eddbf7bb5e7f5e6e7291828ddb5a2ed25dcf6e","aliases":[],"snapshot":{"statement":"A data platform team plans to rename a column in main.sales.orders, one of the most widely reused tables in the lakehouse. Before making the change, the lead engineer must identify every job, query, dashboard, and downstream table that reads from it, without instrumenting any code. What does Unity Catalog provide for this need?","options":["Unity Catalog captures lineage automatically, and Catalog Explorer's Lineage tab shows the table's downstream jobs and dashboards.","Lineage becomes available only after the team enables a dedicated capture job for each table that needs to be tracked.","Lineage tracks Lakeflow Jobs and pipelines but excludes queries run in Databricks SQL and the dashboards built on them.","Unity Catalog offers no lineage capability of its own, so every consuming team must maintain a manual catalog that documents its dependencies."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00172","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b4abc4440074564f04042d9d842c8b0ec8ded2641400c32aa325cb9e4fe3cf23","aliases":[],"snapshot":{"statement":"A compliance officer must produce a list of every principal that ran SELECT against the sensitive table main.hr.compensation during the last 30 days, along with timestamps. The list must come directly from a SQL query in the lakehouse: the team may not download access logs from the cloud provider or export CSV files from the query history UI. Which approach meets the requirement?","options":["Ask the cloud provider for the bucket's access logs and parse the object-level entries that touch the table's files.","Query the system.access.audit system table, filtering the event date and the request parameters that name the table.","Enable query history for each user and export every user's CSV history for offline consolidation into one report.","Ask the account admin to review raw audit events manually, because audit events are visible only to admins and cannot be queried with SQL."],"answer":"B","reference":"https://docs.databricks.com/aws/en/admin/system-tables/audit-logs","referenceTitle":"Audit log system table reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00173","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:1990c7b556f4f4cce1ac06f21f91155fe1bcc428b3ca7b70fad09b914bc7769f","aliases":[],"snapshot":{"statement":"A workspace has years of legacy tables in the workspace-local Hive metastore, still referenced by old jobs using two-part names like legacy_db.orders. The company has now enabled Unity Catalog, and policy requires every production table to fall under its centralized permissions, lineage, and audit.\n\nWhat must happen to legacy_db.orders to meet the policy?","options":["Nothing is needed: enabling Unity Catalog automatically places existing hive_metastore tables under its governance","Run GRANT SELECT through Unity Catalog against the legacy table, which supplies the missing permissions, lineage, and audit trail","Mount the table's underlying storage with DBFS so that Unity Catalog can see and govern the data files directly","Migrate the table into Unity Catalog and update the old jobs to use its three-level name, so it becomes a UC-governed object"],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00174","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9f607121d3d23c83a13657eabe81b770fc86e4828499a388b8abd5dffa4059eb","aliases":[],"snapshot":{"statement":"A company runs five Databricks workspaces, all in the AWS region us-east-1, and wants every workspace to see the same Unity Catalog catalogs, such as main and finance, so that grants and data stay consistent across teams. The platform lead is documenting how the Unity Catalog metastore is organized for this deployment. Which statement is correct?","options":["Unity Catalog uses one metastore per account per region; all workspaces in us-east-1 attach to it and share its catalogs.","Each workspace hosts its own metastore, and administrators can optionally configure replication between the metastores.","Unity Catalog creates one metastore per catalog rather than per region, so the main and finance catalogs would each get their own metastore.","Best practice is to create several metastores in the same region and assign a dedicated one to each workspace."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00175","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:71c4ec2778a77372f11d29a1dfbe22cbdd1fe19833562bcf121f93a58976d717","aliases":[],"snapshot":{"statement":"A nightly Lakeflow Job runs under the service principal loader-sp and writes report files into the Unity Catalog volume main.ops.drop_zone. The job's identity already has USE CATALOG on main and USE SCHEMA on ops. Which grant should an administrator apply so that the pipeline can write its files into the volume?","options":["GRANT SELECT ON VOLUME main.ops.drop_zone TO `loader-sp`, since SELECT is the privilege that controls reads and writes on volumes.","GRANT ALL PRIVILEGES ON CATALOG main TO `loader-sp`, which is the least-privilege grant that still enables the file writes.","GRANT WRITE VOLUME ON VOLUME main.ops.drop_zone TO `loader-sp`, the Unity Catalog privilege that authorizes writing files in a volume.","GRANT nothing at all, because volumes automatically inherit their access rules from the storage credential backing them."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00176","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:9f89ec8277293345fa86524b1ce466f44a3bd988a1e2327f6b9bf1995d13d12f","aliases":[],"snapshot":{"statement":"A new data engineer joins a company whose lakehouse is governed by Unity Catalog. The onboarding documentation consistently refers to tables by fully qualified names such as main.sales.orders, and the engineer asks a senior colleague what the three parts of the name mean. Which explanation is accurate?","options":["main names the metastore, sales is a database, and orders is a table, following the legacy Hive database.table model with a decorative prefix.","main is the catalog, sales is a schema inside it, and orders is a table in that schema, the three-level namespace Unity Catalog enforces.","main names the workspace, sales is the database inside it, and orders is the table, matching the workspace.database.table convention.","main is the cloud region, sales is a storage bucket in that region, and orders is the table stored under that bucket."],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00177","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c8b8043d953da8a2fab699b62463c16d087612cb6e7aec7b99f927ce2f8f2580","aliases":[],"snapshot":{"statement":"A company syncs groups such as data_engineers from its identity provider to Databricks using SCIM provisioning. The governance team wants a single definition of each group, so that a grant issued in any workspace, or directly in Unity Catalog, always applies to the same members. Where should these groups be managed?","options":["Recreate each group manually as a workspace-local group in every single workspace, keeping every member list synchronized by hand across the company.","Keep the groups at the account level, but accept that Unity Catalog data grants cannot reference account groups on securables.","Manage them as account groups synced from the identity provider: they can be used for grants in any workspace, unlike workspace-local groups.","Rely on the SCIM sync to create an equivalent workspace-local copy of each group automatically inside every workspace."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00178","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:8e458aaf364147329adb305922dccc2c0078e162e0263d9739c5f07baba29008","aliases":[],"snapshot":{"statement":"During a quarterly access review, the governance team must list every principal that currently holds any privilege on the Unity Catalog table main.sales.orders, together with the privilege types. The reviewer wants to produce the list with one SQL statement in the query editor. How should the reviewer get the information?","options":["Run SHOW GRANTS ON TABLE main.sales.orders, which lists each principal and the privilege it holds on the table.","Run DESCRIBE TABLE main.sales.orders, because its output includes the current grantees along with the table's columns.","Use the table details page in Catalog Explorer, because grants are visible there and no SQL statement returns them.","Run SELECT * FROM grants, the global view that lists every grant in the metastore without needing to name a specific object."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00179","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:f58fe41ed6f2e4ae5a05bfef1d3ca758f5442501b757ecf86b00e3d3a6ddc0f4","aliases":[],"snapshot":{"statement":"During a security audit, the auditors require separation of duties across the company's Databricks deployment. They ask the platform lead to explain which administrator role manages users, groups, and workspaces at the account level, which role manages an individual workspace, and which role manages catalogs and data in the metastore. Which answer is accurate?","options":["Account admin is the only built-in administrative role; workspace admin and metastore admin are just conventional names that teams use for ordinary groups.","Workspace admin automatically includes metastore admin, so a single role is enough to govern both the workspace and the catalogs in its region.","Account admins manage users, groups, workspaces, and metastores at the account level; workspace admins govern one workspace; metastore admins manage catalogs and data.","All three administrative scopes should be assigned to the same trusted person, which is what the auditors' separation-of-duties requirement expects."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00180","exam":"DBDEA","topics":["platform"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:220efc390c643f51790f5d7467c8eb32a7f7c4ad62f1c79da50ce302a71a96b9","aliases":[],"snapshot":{"statement":"A data engineering team's nightly Lakeflow Job runs on a classic job cluster that spends five to ten minutes launching before the first task starts. The team is evaluating moving this job to serverless compute for jobs. Which statement correctly describes what serverless compute changes for this workload?","options":["Serverless compute has Databricks provision, manage, and scale the compute automatically, so jobs start quickly with no clusters to deploy or configure","Serverless compute still requires the team to create, configure, and maintain a dedicated cluster for the job before each run can start","Serverless compute can only power SQL warehouses, so it is unavailable to Lakeflow Jobs and their notebook tasks","Serverless compute removes the need for Unity Catalog to govern the tables that the job reads from and writes to"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute","referenceTitle":"Compute — Databricks Docs"}}
{"id":"RAN-DBDEA-00181","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d1ca2c096eb99c5090f3a1d48a2d8d3443a6ea78f03baf34b51a47a534a21550","aliases":[],"snapshot":{"statement":"A BI team is choosing a SQL warehouse for analysts who run ad hoc queries and dashboard refreshes with very uneven traffic: long quiet periods followed by bursts of concurrent queries. The team wants compute that starts quickly and scales responsively with that demand. Which warehouse should the team choose?","options":["A classic or pro warehouse, because these provision their clusters faster than serverless warehouses and react to load spikes more quickly","Any warehouse type works, because classic, pro, and serverless warehouses behave identically in startup and scaling and differ only in price","A serverless warehouse, which starts quickly, scales dynamically, and is the type Databricks recommends for most workloads","A serverless warehouse, once the team has selected the instance types that its worker clusters will run on"],"answer":"C","reference":"https://docs.databricks.com/aws/en/compute/sql-warehouse/warehouse-behavior","referenceTitle":"SQL warehouse sizing, scaling, and queuing behavior — Databricks Docs"}}
{"id":"RAN-DBDEA-00182","exam":"DBDEA","topics":["platform"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:51e8b5d555059aee6226d00e4c67f8d700505c36a68ace20f98327f5680d4602","aliases":[],"snapshot":{"statement":"A new analyst asks where each of these artifacts should live in the Databricks workspace: an iterative Python script that will be edited and run cell by cell, a SQL query that must be saved once and rerun every week, and a management-facing page of visualizations. Which guidance is correct?","options":["Use notebooks for all three artifacts, since saved queries and dashboards are not workspace objects and live in separate external tools","Develop Python in notebooks, save the reusable SQL as a query in Databricks SQL, and build the management page as a dashboard","Build dashboards for all three needs, since dashboards fully replace notebooks for developing and running code","Develop Python in notebooks and build dashboards, but keep the weekly SQL in notebooks too, because saved queries cannot be shared"],"answer":"B","reference":"https://docs.databricks.com/aws/en/introduction","referenceTitle":"Introduction to the Databricks Data Intelligence Platform"}}
{"id":"RAN-DBDEA-00183","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:2610c1f7be514fe56a0d30db41a93808beede11078e53852c52251699f4279a3","aliases":[],"snapshot":{"statement":"Finance asks the platform team to break down monthly Databricks compute costs by team and project so that each group can be charged back. The company runs a single workspace whose many clusters and Lakeflow Jobs are shared across teams. Which approach meets this requirement?","options":["Give every team its own workspace whenever possible, because compute costs can only be measured and charged back per workspace","Track costs by cluster name only, since the cluster name is the single identifying field carried into usage reports","Tag only the SQL warehouses, because custom tags are ignored on classic clusters and job clusters","Apply team and project tags to clusters and jobs, and break the tagged usage down in the billing reports"],"answer":"D","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00184","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6c6a7546c867a15e020f076d7f3829dba0d4ba635d753e6d11cdab960ea9b027","aliases":[],"snapshot":{"statement":"A platform team must install a native operating system package on every node of every classic cluster in the workspace before notebooks or job tasks run, and manually configuring each cluster no longer scales as clusters multiply. Which mechanism should the team use to automate this installation?","options":["Init scripts that run on every node at cluster startup, attached per cluster or defined globally for the whole workspace","Manual installation of the package through each cluster's library settings in the UI, repeated by hand for every new cluster","A %pip install cell in the first notebook of every job, which installs the package node-wide for all later sessions","Init scripts configured to run after each task completes, cleaning up the installed packages between runs"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00185","exam":"DBDEA","topics":["platform"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f217e4147832b64d7d65705f33bae0652e9002ca8aadb6825b2c75bfd6b9b36c","aliases":[],"snapshot":{"statement":"A team's latency-sensitive Lakeflow Jobs run on classic job clusters, and every run still loses four to six minutes while cloud instances are acquired and bootstrapped. The team wants to cut that startup delay for its most time-critical nightly runs. Which approach should the team take?","options":["Keep one all-purpose cluster running around the clock and point every job at it, so no job run ever waits for a cluster to start","Rely on an instance pool's pre-warmed instances, which deliver instant cluster startup with no idle-instance charges on the cloud bill","Attach the job clusters to an instance pool, which keeps idle instances ready and cuts startup time at the cost of idle billing","Move the jobs onto a pool-backed SQL warehouse, because instance pools can only accelerate SQL warehouses"],"answer":"C","reference":"https://docs.databricks.com/aws/en/compute","referenceTitle":"Compute — Databricks Docs"}}
{"id":"RAN-DBDEA-00186","exam":"DBDEA","topics":["platform"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ef1b44a50bd97515f6d31d6e932b6f6fab7707028ce0d216ffc3a6e2b3037c17","aliases":[],"snapshot":{"statement":"A newly hired data engineer asks how work is divided when a PySpark notebook runs on a multi-node cluster: where the notebook's code and the Spark session live, and where the actual parallel processing of data happens. Which answer is correct?","options":["All of the work, including data processing, runs on the driver node; the executors are passive and only store the underlying data files","Each executor runs its own full copy of the notebook's code and independently processes its own portion of the data","Driver nodes exist only on high-concurrency shared clusters; single-user clusters and dedicated job clusters run without any driver at all","The driver node hosts the Spark session and coordinates tasks, while executors run the tasks that process data in parallel"],"answer":"D","reference":"https://docs.databricks.com/aws/en/compute","referenceTitle":"Compute — Databricks Docs"}}
{"id":"RAN-DBDEA-00187","exam":"DBDEA","topics":["platform"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:eaf05ed590037c145d33e0089fffd533d4f6626c561d01b306cbcf9c524bf525","aliases":[],"snapshot":{"statement":"A finance analyst reviewing the Databricks invoice asks the platform team what a \"DBU\" actually is, because the bill is expressed in Databricks Units consumed per hour by each compute resource.\n\nWhich explanation is correct?","options":["A fixed dollar amount that is identical across all compute types, so costs vary only with hours run","A unit of processing capability, billed per second at rates that depend on the compute type","A measure of lakehouse storage consumed, charged per gigabyte-month of data retained","A license seat billed per named user, regardless of how much compute that user runs"],"answer":"B","reference":"https://docs.databricks.com/aws/en/compute","referenceTitle":"Compute — Databricks Docs"}}
{"id":"RAN-DBDEA-00188","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:78cb00e8f4a1bd2e401e03799d824374a0bb9041c5405ce0cb8e44bb478eb1dd","aliases":[],"snapshot":{"statement":"A finance tool exports nightly CSV files, each starting with a header row of column names. The team loads them with the statement below, and the first run produced generic columns such as _c0 whose values are header text like transaction_id, with every data row shifted. Which change makes the load treat the first line of each file as a header?\n\n```sql\nCOPY INTO main.finance.payments_raw\nFROM '/Volumes/main/finance/landing/payments/'\nFILEFORMAT = CSV;\n```","options":["Keep the statement unchanged: COPY INTO detects header rows automatically in every supported file format","Add FORMAT_OPTIONS ('header' = 'true') so the reader treats each file's first line as column names instead of data","Run a preprocessing job that removes the first line of every CSV file in the landing volume before each nightly load runs","Change FILEFORMAT to JSON, because CSV files do not accept format options such as header handling"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ingestion/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00189","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:e36fc42cf4d75b9712a4519a366ff6ac25cb674abca04f27c49fc082e90abb41","aliases":[],"snapshot":{"statement":"An insurance data source delivers one complete extract at 02:00 every night, and nothing changes during the day. Analysts refresh their dashboards once each morning, and no downstream consumer needs intraday data. A stakeholder asks why the team is not building a streaming pipeline like a competitor did. What should drive the choice between streaming and batch processing here?","options":["The freshness requirement: daily refresh satisfies every consumer, so scheduled batch is simpler and cheaper; streaming pays off only when data must arrive within minutes","Industry fashion: streaming is always the recommended modern architecture for new data platforms, regardless of any actual latency requirement","Engine unification: because Databricks runs batch and streaming on one unified engine, every ingestion must now be written as a continuous streaming job","Data volume alone: the decision depends exclusively on how many bytes the source produces each night, never on how fresh consumers need the data"],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-engineering/batch-vs-streaming","referenceTitle":"Batch vs. streaming data processing — Databricks Docs"}}
{"id":"RAN-DBDEA-00190","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:43d833bff63d8f7409e05d78298eae662b3019290b0e5b3b0ad5600c7969df45","aliases":[],"snapshot":{"statement":"A supplier landing directory has accumulated more than 40,000 small CSV files over two years of daily deliveries. The nightly COPY INTO run keeps getting slower: most of its time now goes into discovering which files in the path are new and which were already loaded. Total daily volume is modest. What should the team conclude about this slowdown?","options":["No structural limit: COPY INTO handles any file count without degradation, so the slowdown must come from warehouse sizing, not ingestion design","File size only: compacting the thousands of small files into fewer large ones solves everything, because file count never drives ingestion cost","Scale mismatch: COPY INTO suits small to medium sets of files, while Auto Loader is built for this scale, handling millions of files efficiently","Mode mismatch: Auto Loader only serves always-on, real-time streaming, so it cannot replace a nightly batch-style incremental load"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00191","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:cdfa4843dd7410390a254d87309086496cf3db644ff659dd608bf60a01550c72","aliases":[],"snapshot":{"statement":"For months, a nightly COPY INTO has loaded new files from a landing volume into main.ops.bookings. During a schema dispute, an engineer drops and re-creates the table with the correct columns. The next run reloads every file still present in the landing directory, duplicating months of history. Where does COPY INTO keep its record of already-loaded files?","options":["In a workspace-level ingestion service that tracks every loaded file globally, shared by all tables that read from the same cloud storage account","In the source path itself: each landing directory maintains its own tracking metadata, independent of whichever table the data lands in","In the Auto Loader checkpoint directory that backs the job, which should have been backed up before the table was re-created","In the target table's Delta transaction log: dropping and re-creating the table destroyed the loaded-file history, so everything was reloaded"],"answer":"D","reference":"https://docs.databricks.com/aws/en/ingestion/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00192","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b69550a1df843e1f98004c79bd6d8697cb521e1bb1b67f3282b72d737104b8a5","aliases":[],"snapshot":{"statement":"A partner REST API returns orders in pages, and every response includes a next_page cursor. A nightly Lakeflow Job must fetch only orders that are new since the previous run, because reprocessing old pages wastes API quota. The job restarts on failures and its compute terminates after each run. How should the ingestion position carry over between runs?","options":["Replay from the first page every night and let a MERGE against the target re-deduplicate the whole order history again on every run","Persist the newest cursor in a small Delta control table, read it at start, and pass it to the next run as a job parameter","Write the cursor to a file on the job cluster's driver disk, which survives across runs even though the cluster itself terminates","Store nothing: the partner API remembers each client's last position on the server side, so the job can simply ask for whatever orders are new"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00193","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:ac2268248b38bd8bd932b03acdd9b0197a5d6e53bc3b1e4bbecde902e69f9afb","aliases":[],"snapshot":{"statement":"A nightly job pages through a partner events API that occasionally times out. The job's retry logic can re-request a page, so the same event_id sometimes arrives twice in one night; the API cannot promise exactly-once delivery. Each event must appear exactly once in the target Unity Catalog table. Which design makes this ingestion reliable?","options":["Pause the pipeline until the partner can guarantee exactly-once delivery for every event_id, before writing anything into the table","Append the raw responses with plain INSERT INTO and require analysts to filter duplicate event_ids out of every downstream dashboard and report","Accept at-least-once delivery and write idempotently: MERGE INTO the target on event_id, updating matches and inserting rows that are new","Disable the job's retry logic so a timed-out call is never repeated, accepting that lost responses mean permanently missing events"],"answer":"C","reference":"https://docs.databricks.com/aws/en/delta/merge","referenceTitle":"Merge data (Delta Lake upsert) — Databricks Docs"}}
{"id":"RAN-DBDEA-00194","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:df2e5f0c47e0b6d6190f557a0d74987211ffcbd362d153cff685131ca7435879","aliases":[],"snapshot":{"statement":"A data platform team is cataloguing the ingestion options available through Lakeflow Connect for an architecture review. One column of their catalogue must record who operates each connector: they want zero infrastructure to run themselves for standard enterprise sources, while accepting that some niche sources will run on the team's own compute.\n\nWhich classification matches how Lakeflow Connect organizes its connectors?","options":["Managed connectors are Databricks-hosted and operated; community connectors are open-source packages the team runs itself","Every connector in the Lakeflow Connect catalog is hosted and operated by Databricks on serverless compute, so the team runs no infrastructure","Community connectors are operated by Databricks on the team's behalf at no additional charge","The catalog covers only databases and object storage; SaaS and streaming sources need custom code"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/lakeflow-connect/","referenceTitle":"Lakeflow Connect connector concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00195","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:25b0abf1d1f575d444d3ac8dc64f3e20ad1ddf8f559bf8db93bbda3f2c924712","aliases":[],"snapshot":{"statement":"A Structured Streaming query reads a 12-partition Kafka topic that feeds a Delta table. During traffic spikes, single micro-batches grow enormous and overwhelm the cluster. The team wants to cap how much data each micro-batch consumes and asks what controls the number of concurrent read tasks. Which configuration answers both concerns?","options":["Executor count: read parallelism depends only on how many workers the cluster has, and the topic's partition layout plays no role","Topic repartitioning: changing the partition layout is the only throttling mechanism, because Spark exposes no rate-limiting option for Kafka","Manual pacing: Kafka sources cannot be rate-limited in Spark at all, so an operator must pause and resume the query during spikes","Offset cap: set kafka.maxOffsetsPerTrigger to bound each micro-batch; read tasks follow the topic's partitions, raised by minPartitions"],"answer":"D","reference":"https://docs.databricks.com/aws/en/connect/streaming/kafka/","referenceTitle":"Connect to Apache Kafka — Databricks Docs"}}
{"id":"RAN-DBDEA-00196","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0d99e879665ba324c6abadd2f33f91c9b33aeb841aa1b5a1c4eea68a04608c87","aliases":[],"snapshot":{"statement":"A bronze streaming table in a Lakeflow pipeline ingests device events. Rows with a null device_id arrive occasionally; they must never reach the table, but the team wants the failure counts visible in pipeline metrics. The current definition is:\n\n```sql\nCREATE OR REFRESH STREAMING TABLE main.bronze.device_events\nAS SELECT * FROM STREAM read_files('/Volumes/main/bronze/devices/');\n```\n\nHow should the pipeline handle the invalid rows?","options":["Add an expectation with the default warn action: invalid rows still land in the table, and only failure counts appear as metrics","Add a CONSTRAINT ... EXPECT (device_id IS NOT NULL) ON VIOLATION DROP ROW clause: invalid rows are dropped and their count is logged","Leave the definition as is and filter WHERE device_id IS NOT NULL in every downstream dashboard and query that reads the table","Direct all raw events into a separate quarantine table and have an on-call engineer delete the invalid rows manually after every pipeline refresh"],"answer":"B","reference":"https://docs.databricks.com/aws/en/ldp/concepts/streaming-tables","referenceTitle":"Streaming tables (Lakeflow pipelines) — Databricks Docs"}}
{"id":"RAN-DBDEA-00197","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0f2d2d9658b58dbf354fa077ffb4063638ea6183981d720e944d841102d6c5c0","aliases":[],"snapshot":{"statement":"A producer team can deliver its event feed either as Avro or as raw XML files, and it asks the ingestion team which formats can be loaded directly into Delta tables with COPY INTO or Auto Loader — without any custom parsing code — before they commit to a serialization.\n\nWhich answer is accurate for the current platform?","options":["Only CSV and JSON load natively; Avro and XML must both be converted to Parquet before ingestion","Avro requires a Lakeflow Connect database connector, while XML loads directly with COPY INTO","Both load natively: formats include CSV, JSON, XML, Avro, ORC, Parquet, text, and binary files","Auto Loader reads XML, but COPY INTO is limited to CSV, JSON, and Parquet sources"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/copy-into","referenceTitle":"Get started using COPY INTO — Databricks Docs"}}
{"id":"RAN-DBDEA-00198","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9a8ef1107eaf3cf3a4ae8bc13f6b550cc6a0cb9997bc8b8d64958de570980c63","aliases":[],"snapshot":{"statement":"Files land in a Unity Catalog volume at unpredictable hours: some days at 08:00, some at 23:00, and some days not at all. Consumers want each delivery loaded shortly after it arrives, but the team rejects an always-on stream and refuses fixed-interval scans that mostly find nothing. Which design ingests every delivery soon after it lands?","options":["A Lakeflow Job with a file arrival trigger on the volume, running Auto Loader with Trigger.AvailableNow: each run wakes on arrival, processes pending files, and stops","An always-on stream with trigger(processingTime='1 minute') on continuous compute, so a micro-batch checks for new files every minute around the clock","A cron schedule firing every five minutes that runs COPY INTO over the whole directory, accepting that most runs find nothing new","A scheduled job whose Trigger.AvailableNow step runs alongside a parallel always-on stream, because AvailableNow requires an active query to work"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/file-arrival-triggers","referenceTitle":"Trigger jobs when new files arrive — Databricks Docs"}}
{"id":"RAN-DBDEA-00199","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:55f50b2edd06d2ea2558ee2b4a3d42df412d05a18cfab78fd19dba9d02c9c308","aliases":[],"snapshot":{"statement":"Two teams each start their own Auto Loader stream over the same landing directory, writing to the same Delta table, main.bronze.clicks. Each stream has its own checkpoint location and its own Lakeflow Job. Days later, the table holds exactly two copies of every ingested file. Why did the duplicates appear, and what is the correct fix?","options":["Delta Lake should have detected the distinct checkpoints and deduplicated the writes; the duplicates point to a corrupted transaction log instead","Auto Loader refuses a second stream targeting the same table with a lock error, so one of the jobs must have silently skipped its files","Each stream tracks its own progress, so both discovered and ingested the same files; run a single stream per table or merge the sources","The producer uploaded every file twice into the landing directory, because separate stream writers can never create duplicates by design"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/best-practices","referenceTitle":"Auto Loader best practices — Databricks Docs"}}
{"id":"RAN-DBDEA-00200","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:a3066abeea392be2d8e543fe5e32fb34cca032e2b90e25872872cf10c3081d2a","aliases":[],"snapshot":{"statement":"A landing bucket holds three years of historical files that must be loaded into a new table exactly once by a dedicated backfill run. After that run completes, ongoing arrivals must be captured incrementally by an Auto Loader stream, without reprocessing a single historical file and without backfill flags in the streaming code.\n\nWhich design meets both requirements?","options":["Point the Auto Loader stream at the whole bucket with a low cloudFiles.maxFilesPerTrigger and let it chew through history first","Load history with COPY INTO and keep using COPY INTO nightly, because Auto Loader would reload everything anyway","Run one backfill batch with COPY INTO over the history, then start Auto Loader with cloudFiles.includeExistingFiles set to false","Start the stream first and run the backfill afterwards, since Auto Loader skips files loaded by other tools"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00201","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:04dbcacbe821d70b8f8f9d70a80b2cdbb0099e7ab2e951901688ecdf9b0e5e17","aliases":[],"snapshot":{"statement":"An engineer with a background in classic Kafka consumers is setting up a Structured Streaming read from a Confluent Kafka topic and asks where the reading position — which offsets have already been consumed — is maintained for the stream, since that determines what happens after a restart.\n\nWhat should the team tell them?","options":["In the query's checkpoint on durable storage: offsets are committed every batch, so restarts resume from it","In the broker's consumer-group offsets topic, exactly like an application using the classic Kafka consumer","In memory on the driver, which means every restart begins from the earliest offset unless configured otherwise","In a Databricks-managed table that the team must inspect and repair manually after failures"],"answer":"A","reference":"https://docs.databricks.com/aws/en/connect/streaming/kafka/","referenceTitle":"Structured Streaming from Kafka — Databricks Docs"}}
{"id":"RAN-DBDEA-00202","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a0fb4d0c0dc7366b6f31fe9634fede054154a0154e149fdac84a51f7263415cd","aliases":[],"snapshot":{"statement":"During an ingestion review, the team lists their sources: PostgreSQL, Salesforce, Confluent Kafka, Google Analytics, and a niche SaaS recruiting tool with no Databricks-provided connector. For that last source, they are evaluating vendors from the Databricks partner ecosystem rather than writing custom code.\n\nWhat role do partner connectors play here?","options":["Partner connectors replace Unity Catalog governance for the sources they ingest, so no grants apply to their tables","Partner connectors are required for every SaaS source, including well-covered ones like Salesforce","Partner connectors extend coverage to sources without a managed connector, through established ecosystem vendors","Partner connectors run only inside Lakeflow Jobs, executing as notebook tasks within the job graph"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/lakeflow-connect/","referenceTitle":"Lakeflow Connect connector concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00203","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:474f9e43289cce5e9fb471ad5b8cfc659b318730d4026636538195b5869f7469","aliases":[],"snapshot":{"statement":"Alongside the tabular feeds, a document-processing pipeline receives scanned invoices, signature images, and PDF contracts that downstream jobs will process but that will never be queried as rows and columns. The governance program still requires this data to live under Unity Catalog access control.\n\nWhere should these non-tabular files be stored?","options":["As binary blobs embedded in a Delta table column, with one row created per document","In a Unity Catalog volume: /Volumes paths are governed by volume privileges","In the workspace's DBFS root, which Unity Catalog also governs by default","They cannot be governed at all, because Unity Catalog only manages tables"],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00204","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:8c61cc3a9bedcf5e631aadea79f63f2cde879fff9d90ab00aa7c4df9a291b525","aliases":[],"snapshot":{"statement":"A business analyst has produced a one-off CSV extract on their laptop and needs it available to the whole team for a shared analysis. The team's storage is fully governed by Unity Catalog, the analyst has no cloud console access, and emailing the file around is against policy.\n\nHow should the file reach the team?","options":["Email the CSV to the data engineering team, who each load it locally into their own sessions","Attach the file to a notebook's revision history so teammates can check it out","Push the CSV to a public cloud location and share the URL with the team","Upload it via Catalog Explorer into a Unity Catalog volume and share the governed path"],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00205","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d3784d42257757d7f78d113afd7b80cb3b8933b1d952fb9cd17bc4ab07e94724","aliases":[],"snapshot":{"statement":"A feed has a strict schema contract with the producing team: the ingested table must contain exactly the agreed columns, and anything the producer adds outside the contract must be captured for review rather than silently appearing in the table. The ingestion uses Auto Loader.\n\nHow should the stream be configured?","options":["Supply the agreed schema explicitly with evolution off; unexpected columns then land in the rescued data column","Use schema inference with default evolution, so new columns appear immediately and visibly in the table for review","Let inference run without a schema location, which freezes the schema on the first file read","Set evolutionMode to addNewColumns and review the table's history after each run for surprises"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00206","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:01ce359d3b097fbf6c05d9491cc07c021e35158cd3d3942832ac0ba0219635f6","aliases":[],"snapshot":{"statement":"A managed database connector will replicate data from a production PostgreSQL instance into Unity Catalog. Of the roughly forty schemas in the instance, the team is only allowed to replicate twelve specific tables, and those tables must land in the ops schema of the main catalog — not wherever a default would put them.\n\nWhat must the team configure?","options":["The connector's table selection (the twelve approved tables) plus the destination catalog and schema for the replicas","Nothing beyond credentials: a managed connector replicates the full database by definition, with a fixed destination","The source table list only, because Unity Catalog always chooses destinations automatically per connector","A separate workspace per replicated table, so each lands under its own governance boundary"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/lakeflow-connect/","referenceTitle":"Lakeflow Connect connector concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00207","exam":"DBDEA","topics":["ingestion"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:46579a53ef8c5528daca1226288ceceaa4bf87e4c4b10abeb2155ebc2481230f","aliases":[],"snapshot":{"statement":"A nightly JDBC extraction reads a 200-million-row table from an external database and then filters it down to the last three days inside Spark. Network costs are enormous, and the external database is barely coping with the full scans.\n\nWhich change pushes the filtering work to the source database?","options":["Increase numPartitions so the extraction parallelizes across more JDBC connections and the full scan finishes faster","Wrap the predicate into the dbtable subquery, so the source database applies it before returning rows","Cache the DataFrame before filtering, so the second run skips the JDBC read","Convert the read to Auto Loader and let it discover only the new rows"],"answer":"B","reference":"https://docs.databricks.com/aws/en/archive/connectors/jdbc","referenceTitle":"Query databases using JDBC — Databricks Docs"}}
{"id":"RAN-DBDEA-00208","exam":"DBDEA","topics":["ingestion"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:10d64181ae3e0744ce1fcf69c5409e0b98ae194b4a3864c4260173547aa1d9b0","aliases":[],"snapshot":{"statement":"A data provider delivers files inside a Hive-style directory layout — year=2026/month=10/day=27/part-001.json — and the ingestion team wants the year, month, and day values to become real columns of the target table without parsing file paths by hand.\n\nWhich Auto Loader capability provides this?","options":["pathGlobFilter, with a glob pattern per date component applied once per ingestion","A UDF that splits the input path column on the equals sign and the year, month, and day tokens in every file","Partition inference: Auto Loader derives partition columns from the directory layout of the files","It is not possible, because directory layout information is discarded during ingestion"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ingestion/cloud-object-storage/auto-loader/","referenceTitle":"What is Auto Loader? — Databricks Docs"}}
{"id":"RAN-DBDEA-00209","exam":"DBDEA","topics":["ingestion"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6f930f94de3739c1f6742750089711a45748a1e7041c168711aa530b7c1f146a","aliases":[],"snapshot":{"statement":"A producing system can publish its events either as a continuous stream to a Kafka topic or as files dropped into cloud storage every few minutes. The consuming lakehouse team must choose one path, and both would technically work.\n\nWhich consideration should drive the choice?","options":["Requirements decide: Kafka for low-latency continuous streams; file drops for simpler operation and reprocessing","Kafka should always be preferred, because it is the more modern protocol for publishing and moving event data","File-based ingestion cannot be incremental and always reloads the full history on every single run","The choice has no downstream effect, so either path is operationally equivalent"],"answer":"A","reference":"https://docs.databricks.com/aws/en/ingestion/lakeflow-connect/","referenceTitle":"Lakeflow Connect connector concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00210","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:631fe3360bdef546c9b71ebe3cc8bde208cc95854b313c371f259899a1844c15","aliases":[],"snapshot":{"statement":"A nightly sales report joins catalog.sales.facts_sales to catalog.sales.dim_stores on store_id to append each store's region and manager. After the join, the report is missing thousands of sales: several new stores started selling before their dim_stores records were created. The business rule requires every sale to appear in the report even when store attributes are unknown.\n\nWhich join change satisfies this rule?","options":["Switch to a LEFT JOIN cautiously — an outer join still discards non-matching rows before combining, so the unregistered stores would keep disappearing","Replace the INNER JOIN with a LEFT JOIN — sales whose store_id has no dimension row are kept, with NULLs filling the store attribute columns","No join change is needed — Spark inspects the table sizes and decides on its own whether INNER or LEFT semantics apply to the query","Keep the INNER JOIN and add an ORDER BY — INNER and LEFT return exactly the same rows and differ only in how the output is sorted"],"answer":"B","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00211","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e34ebbee72c3b4fbf864ddd65cc10d2a2582f8cf8a653c8eecb1b768182821f5","aliases":[],"snapshot":{"statement":"A daily consolidation reads two regional extracts into DataFrames. Both extracts carry the same four columns — order_id, order_date, amount, region — but the EU file delivers them in a different column order than the US file. The combined output must pair every column with its correct data regardless of the position in which each file lists it.\n\nHow should the engineer stack the two DataFrames?","options":["eu.union(us) — union() resolves columns by name automatically whenever the two schemas differ in column position","Reorder one DataFrame with an explicit select() first — manually aligning the column positions is the only safe way to combine the extracts","It cannot be done — Spark refuses to combine DataFrames whose columns are not declared in the same order","eu.unionByName(us) — it pairs columns by name between the two DataFrames, so position differences cannot mismatch the data"],"answer":"D","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00212","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:3344a953a7a55af5ee7ca0933a265c46b0cc7c112303d61c9ae3cee624a2f846","aliases":[],"snapshot":{"statement":"Yesterday, the query plan for the nightly reconciliation broadcast a 30 MB currency dimension next to a multi-terabyte fact table automatically — no hint appears in the query. Today the same pattern runs as a shuffle-heavy sort-merge join, because the platform team lowered the size threshold that governs automatic broadcasts of small tables in joins.\n\nWhich configuration parameter sets that threshold?","options":["spark.sql.autoBroadcastJoinThreshold — the size limit at or below which the planner broadcasts a join side automatically, with no hint required","No parameter is involved — automatic broadcast joins only happen when the query text includes an explicit broadcast() join hint","The broadcast threshold is a fixed engine constant — a platform team cannot raise or lower it through configuration","spark.sql.shufflePartitions — its value determines how large a table must be before the planner stops broadcasting it"],"answer":"A","reference":"https://docs.databricks.com/aws/en/optimizations/aqe","referenceTitle":"Adaptive Query Execution — Databricks Docs"}}
{"id":"RAN-DBDEA-00213","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:9dbb643a94f48b1c559aca9ca4ddae9438f58ade051175940382a954d9ab0b24","aliases":[],"snapshot":{"statement":"In an exploratory notebook, an engineer wants to inspect catalog.events.bronze_events, a table with roughly 200 million rows. She runs the cell below, and a few minutes later the run fails with an out-of-memory error raised on the driver process.\n\n```python\nrows = spark.read.table(\"catalog.events.bronze_events\").collect()\n```\n\nWhy did this fail, and what is the correct practice?","options":["collect() fans the rows out across the executors for inspection — the driver out-of-memory error must have a different cause, such as network buffers","Prepend coalesce(1) — collapsing the DataFrame into a single partition is what allows collect() to return without exhausting driver memory","collect() ships every row into the driver's own memory — inspect with display(), take(n), or limit(), and keep full-scale processing on the cluster","The driver never holds data in Spark — only executors allocate memory, so the failure proves the executors were undersized for the table"],"answer":"C","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00214","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:689306990fa55145f8c8faa2930864e0cceae5076ab8b8c7fda74c4ea9ae375f","aliases":[],"snapshot":{"statement":"A gold aggregate — weekly revenue by region — is currently a plain Delta table that a nightly Lakeflow Job fully rewrites with INSERT OVERWRITE. The team is evaluating replacing it with a materialized view defined in a Lakeflow Declarative Pipeline, and the lead must explain the operational difference between the two approaches.\n\nWhich comparison is correct?","options":["A materialized view still requires the team to write, schedule, and maintain the refresh job that keeps its cached results up to date","A plain Delta table refreshes itself on a schedule it owns, so wrapping it in a nightly Lakeflow Job rewrites the same data twice","A materialized view is declarative — the platform refreshes it on its configured schedule, incrementally processing source changes; a table plus job keeps refresh control with the team","Materialized views and Delta tables are the same storage object — the two names only reflect whether a pipeline created them"],"answer":"C","reference":"https://docs.databricks.com/aws/en/ldp/concepts/materialized-views","referenceTitle":"Materialized views (Lakeflow pipelines) — Databricks Docs"}}
{"id":"RAN-DBDEA-00215","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:39316db328c1cb93a32c1836819fd101e169d9de53c0eeb8fa7ea279dfb99caa","aliases":[],"snapshot":{"statement":"The source system resent yesterday's events file, and the bronze table now contains rows that are exactly identical across every column — a full-row duplicate of each event. The silver table must contain exactly one row per event, and for this feed a duplicate means an exact copy of the whole row.\n\nWhich PySpark operation removes these duplicated rows?","options":["dropDuplicates() — it removes rows that are identical across all columns; dropDuplicates(['event_id']) is the variant that keeps one row per key","dropDuplicates() cannot run without a column list — Spark requires the deduplication key to be specified explicitly in every invocation","distinct() — it inspects the schema and picks an arbitrary business key, such as event_id, to deduplicate the rows on its own","A row_number() window ranked over all columns — exact duplicates can only be removed by ranking the identical rows and filtering down to rank one"],"answer":"A","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00216","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:52f6c870426b24c8d079d83b790c71ca9c9129383ffc92215ad33e00f3ecf47e","aliases":[],"snapshot":{"statement":"A monthly revenue report reads from a table with one row per month and region: the twelve months are values in a month column. The dashboard specification now requires one row per region with a separate output column for each month — Jan, Feb, and so on — each holding that month's revenue total.\n\nWhich SQL capability produces this shape directly?","options":["Hand-written conditional aggregation — twelve CASE WHEN expressions are the only way to reshape rows into columns in SQL, one per month","PIVOT — it turns each distinct value of the month column into its own output column, with the aggregation applied in the same statement","UNPIVOT — it rotates the month rows into columns and computes the aggregates while transposing the table","explode — it expands the month values across the width of the result set, giving one column per month in a single pass"],"answer":"B","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00217","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:d315a2cc82e9e624bec9af6d02225f7a8f091ae37f9179166f360ed5e91c2e63","aliases":[],"snapshot":{"statement":"A gold orders table feeds a dashboard that averages the rating column, and stakeholders want missing ratings to count as 0 instead of NULL. The engineer preparing the aggregation must keep every order row in the output while making each NULL rating read as 0.\n\nWhich PySpark operation achieves this?","options":["dropna() on rating — removing the rows where rating is NULL keeps the dashboard free of missing values entirely","fillna(0) on rating — it drops the rows whose rating is NULL, so only genuinely rated orders remain in the output","Registering a Python UDF — no built-in DataFrame operation can substitute a constant for NULL values in a column","fillna(0) on rating — it replaces each NULL rating with 0 while keeping every order row in the DataFrame"],"answer":"D","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00218","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:fe09ac12761d96f60b44a3ea030e72f2adf745b23819cc968aceb1f691a91442","aliases":[],"snapshot":{"statement":"Phone numbers arrive in the bronze table in five different formats — '(11) 98888-7777', '5511988887777', '+55 11 98888-7777', among others. The silver contract requires every number normalized to E.164, such as +5511988887777. The transformation must run as native Spark expressions in the pipeline, with no custom code registered on the cluster.\n\nWhich function fits this requirement?","options":["A Python UDF — regex-based reformatting always requires custom code registered on the cluster before the pipeline can run it","regexp_extract() — Spark's regex functions can only pull matching groups out of a string; rewriting the full value is not supported","regexp_replace() — it matches a pattern with capture groups and rebuilds the string in the normalized form in one native call","split() with a limit — cutting on every non-digit character resolves any phone format in a single pass of the expression"],"answer":"C","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00219","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b46669c315668796f5b0919b3e5a98f2a0f2d98e71fb56e73e507d127709e5a7","aliases":[],"snapshot":{"statement":"A bronze table lands the amount column as STRING, with valid decimals like '125.50' and occasional junk like 'N/A'. The silver job must produce a DECIMAL(10,2) column, and every source row must survive: failed conversions should become NULL so the row can be routed to quarantine. Yesterday's run crashed with a CAST_INVALID_INPUT error on the first 'N/A' value.\n\nWhich function meets the requirement?","options":["try_cast(amount AS DECIMAL(10,2)) — it returns NULL for values it cannot convert, preserving the row for quarantine routing","cast(amount AS DECIMAL(10,2)) — Spark substitutes a default 0 whenever the string is not a valid decimal number","rlike() pre-validation — a regex can confirm the format but cannot produce the typed DECIMAL column the silver contract requires","No change is needed — Spark's cast is already lenient and never raises an error when the input string is malformed"],"answer":"A","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00220","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:3c3aa7ea8d99fe0e580e5186af7e5697508faf99e6e2ac145df6d53a5e7a6c9e","aliases":[],"snapshot":{"statement":"Compliance requires that every payment amount be non-negative. An engineer runs the statement below on a Delta table that already holds millions of historical rows, including negative refunds imported from a legacy migration:\n\n```sql\nALTER TABLE catalog.finance.payments\n  ADD CONSTRAINT chk_amount CHECK (amount >= 0);\n```\n\nWhat is the outcome of this statement, and why?","options":["It succeeds — the CHECK applies only to rows written after it is added, and existing rows are never re-validated against it","It fails with a parse error — CHECK constraints are not supported on Delta tables, where only NOT NULL can be enforced","It succeeds — the constraint is created in a disabled, logging-only state and only starts validating once it is enabled","It fails — ADD CONSTRAINT verifies every existing row against the CHECK, and the negative rows block it until they are corrected or removed"],"answer":"D","reference":"https://docs.databricks.com/aws/en/tables/constraints","referenceTitle":"Table constraints — Databricks Docs"}}
{"id":"RAN-DBDEA-00221","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ea3f3fe891945d5e8cca489f0809205022dee91ebcd75794e51a35a25740e4b8","aliases":[],"snapshot":{"statement":"A new data architect is reviewing the team's medallion pipeline and asks the engineers to state what belongs in the silver layer and what belongs in the gold layer. The team must agree on the canonical responsibility of each layer before designing any new gold tables.\n\nWhich statement describes the two layers correctly?","options":["Gold stores the raw records exactly as extracted from the source systems, preserving full source fidelity for audit and future reprocessing","Silver holds cleaned, validated, conformed data, one row per business entity; gold holds the aggregates and metrics that reports consume","Silver is the layer that serves dashboards and BI tools directly, while gold is where newly landed data first arrives for ingestion","The layers are interchangeable copies of the same data — only naming conventions distinguish one from the other"],"answer":"B","reference":"https://docs.databricks.com/aws/en/lakehouse/medallion","referenceTitle":"Medallion lakehouse architecture — Databricks Docs"}}
{"id":"RAN-DBDEA-00222","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:e82d3d9cc75965ab3a4273c0a4c66e1610d4b735045d43f7818423f5605b7ff1","aliases":[],"snapshot":{"statement":"A silver pipeline must process only the rows that changed in a bronze Delta table since its last run — inserts, updates, and deletes alike — instead of rescanning the whole bronze table every night.\n\nWhich mechanism provides the incremental change stream?","options":["Rescan the bronze table nightly and MERGE on the primary key, accepting the full read every time","Read the bronze table as a streaming source, which surfaces updates and deletes as well as inserts","Enable change data feed and read the changes with table_changes, which returns each change with its type","Keep a hand-built control table of processed ids, refreshed by a nightly full scan"],"answer":"C","reference":"https://docs.databricks.com/aws/en/tables/features/change-data-feed","referenceTitle":"Use change data feed — Databricks Docs"}}
{"id":"RAN-DBDEA-00223","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:48e2e7529f5928e0f0d5dfabef6cdc1f20db90da4a6d721576392cd89cef77e9","aliases":[],"snapshot":{"statement":"A bronze events table stores customer details in a nested column, declared as follows:\n\n```sql\ncustomer STRUCT<name: STRING, address: STRUCT<city: STRING, country: STRING>>\n```\n\nThe silver table must expose the customer's city as a top-level column. The engineer wants the simplest native way to read that nested field, without flattening or custom code.\n\nWhich approach is correct?","options":["Dot notation — select customer.address.city, which resolves nested struct fields natively in SQL and PySpark; arrays use bracket indexing like items[0]","Explode the structs first — a nested field can only be read after the entire struct has been flattened into separate rows of the DataFrame","Rewrite the transformation in SQL — PySpark DataFrames cannot traverse fields that live inside a struct column","Register a Python UDF — native functions cannot reach inside a struct, so the field must be parsed out in custom code"],"answer":"A","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00224","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:84e3130076856dfd3711fd1910af256d301cacb8775201c27875d3905860dd7e","aliases":[],"snapshot":{"statement":"A data engineer builds an intermediate DataFrame in her notebook: a read of a large Parquet dataset from cloud storage, a join with a 5 GB dimension, a filter to active regions, and a deduplication pass. Five cells later in the same notebook reuse that DataFrame, and each one currently re-runs the whole lineage, taking about six minutes per cell. She wants the reuse to be fast and the memory released when the notebook ends.\n\nWhich approach fits?","options":["Call cache() before the first action runs, so that even the initial computation that materializes the DataFrame becomes faster","Call cache() in the very first cell and rely on the cached data surviving cluster termination, so that tomorrow's job run reuses it","Do nothing, because Spark automatically caches every DataFrame in memory after its first action and reuses it for free","Call cache() on the DataFrame, materialize it with an action once, and call unpersist() when the last consuming cell finishes"],"answer":"D","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00225","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c09906d1b7fd4ef895bccb9a8ff56ae48475b592ef5e437d5e4fa486dbacb8f6","aliases":[],"snapshot":{"statement":"Marketing asks the data team for a very specific list: every registered customer who did not place a single order in October. The relevant tables are catalog.sales.customers, with one row per customer, and catalog.sales.orders, with many rows per order. The engineer wants a single PySpark join call that returns exactly those customers and nothing else.\n\nWhich join type should the engineer use?","options":["'left_semi', which returns the customers whose key has no match among the orders","'left_anti', which returns the customers whose key has no match among the orders","No native join type can do this; it must be emulated with a LEFT JOIN plus a filter for NULLs on the right side","'left_anti', which deletes the matching order rows from the orders table so that the customers appear unmatched"],"answer":"B","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00226","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:eb33ec0c26caafd8a34b6c88282fe9acd654d1467d335529498836146f1b8cd8","aliases":[],"snapshot":{"statement":"A re-delivered bronze file duplicated customer records: the same customer_id now appears up to three times, and the copies differ in their updated_at values, so no two rows are fully identical. For the silver table, the team must end up with exactly one row per customer_id, and keeping any one of the copies is acceptable. A teammate proposes calling distinct() on the bronze DataFrame.\n\nWhat is the difference between the two candidate calls?","options":["distinct() compares every column, so rows that differ in updated_at all survive; dropDuplicates(['customer_id']) keeps a single row per key","distinct() and dropDuplicates() are interchangeable, so the teammate's call would already deduplicate on the customer_id key","distinct() accepts a column list, as in distinct('customer_id'), and deduplicates on just the columns passed to it","dropDuplicates(['customer_id']) cannot work by itself; it must be combined with a row_number() window to pick the surviving row"],"answer":"A","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00227","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:08575ceaeb940ef5ecfd907eb4e4fd7e068b94ca961164b94f08097dd0edc927","aliases":[],"snapshot":{"statement":"An analyst joins customers to loyalty_accounts on customer_id and is surprised to see that customers whose customer_id is NULL disappear from the output. The join call passes no join type:\n\n```python\ncustomers.join(loyalty_accounts, \"customer_id\")\n```\n\nA teammate argues that switching to a left join will not help, because a NULL key never matches anything anyway. Which statement is correct?","options":["NULL is treated as equal to NULL in the join, so those customers did match loyalty rows that also carry a NULL key","A join whose key column contains NULL fails outright with an error, and those rows were rejected while all the other rows joined normally","NULL never satisfies an equality comparison, so the default inner join drops those rows — a left join would keep them, padded with NULLs","Spark converts NULL keys to empty strings before joining, so the customers survive in the output under a different value"],"answer":"C","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00228","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c4e1cee8ff60aaa5efa28d2e26e3fd97f5d56caeb75ee2de7fb2f5b9042d663c","aliases":[],"snapshot":{"statement":"A monthly revenue report must show one row per month with two separate columns, web_revenue and store_revenue. The source is a large orders table with the columns event_month, channel, and amount, and the report must compute both totals in a single pass over the table.\n\nWhich construct computes the two metrics in one aggregation?","options":["One query per channel, each filtering WHERE channel = 'web' or WHERE channel = 'store', whose result sets are combined into the report afterwards","GROUP BY channel with SUM(amount), which yields one row per channel carrying that channel's revenue total","PIVOT on channel, which is limited to count aggregations and therefore cannot produce revenue sums per channel","SUM(amount) FILTER (WHERE channel = 'web') and SUM(amount) FILTER (WHERE channel = 'store') within one GROUP BY event_month"],"answer":"D","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00229","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c7ff8cc3dc26915adbd53c5a51c59eee39945d3b3ccd8888f10305e1fb3709e7","aliases":[],"snapshot":{"statement":"The bronze orders table stores an ARRAY<STRING> column named coupon_codes. The attribution notebook needs one row per order_id and coupon pair, but many orders carry no coupons at all — their array is empty or NULL — and those orders must still appear in the output, with NULL in the coupon position, so that order totals reconcile.\n\nWhich expression should the notebook use?","options":["explode(coupon_codes) — rows whose array is empty or NULL are preserved, with NULL in the generated column","explode_outer(coupon_codes) — rows whose array is empty or NULL are preserved, with NULL in the generated column","posexplode(coupon_codes) — as the positional variant of explode, it is the one that preserves rows with empty or NULL arrays","No generator is safe here — an empty array aborts the query, so orders without coupons must be filtered out before exploding"],"answer":"B","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00230","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:599a1c55ea4a68884539d1ab780436e89489c968497db87be8beeac233e02e9e","aliases":[],"snapshot":{"statement":"A reporting query must return every order, joined to its payment when one exists, while showing only approved payment details:\n\n```sql\nSELECT o.order_id, p.status\nFROM orders o\nLEFT JOIN payments p ON o.order_id = p.order_id\nWHERE p.status = 'APPROVED';\n```\n\nAfter running it, the analyst notices that orders without any payment are missing from the results. Why did they disappear, and what is the fix?","options":["The WHERE clause runs after the join: NULL-padded orders fail p.status = 'APPROVED', turning the LEFT JOIN into an inner join — move the predicate to ON","ON and WHERE are interchangeable in outer joins, so the missing orders must have been lost by a different filter somewhere upstream","Spark automatically rewrites WHERE predicates on the right table into the ON clause, so rewriting the query manually is pointless","A LEFT JOIN by definition never returns orders that lack a matching payment; preserving them requires a FULL OUTER JOIN instead"],"answer":"A","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00231","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:4673fdbcbed2683520b22acab9c791f216e914a9b63efe2dda85a6d378db89d2","aliases":[],"snapshot":{"statement":"The weekly engagement report counts reviews with the expression SELECT count(comment_text) FROM catalog.reviews. The reported number comes out lower than the table's actual row count, and QA confirms that many customers submit a star rating only, leaving comment_text NULL. The report must count every review.\n\nWhich expression and explanation are correct?","options":["count(comment_text) counts every row, NULLs included, so the missing reviews must be dropped somewhere else in the pipeline","count(*) skips any row that contains a NULL in one of its columns, so both expressions under-report on nullable tables","count(expr) counts only rows where expr is not NULL; count(*) counts every row, so the report needs count(*) to total all reviews","count(*) and count(col) are equivalent and always return the same number, so the shortfall must come from a stale result cache"],"answer":"C","reference":"https://docs.databricks.com/aws/en/sql/language-manual/sql-ref-functions","referenceTitle":"Spark SQL built-in functions — Databricks Docs"}}
{"id":"RAN-DBDEA-00232","exam":"DBDEA","topics":["transformation"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:9c2ae1fb8f8799c75b5e24fc9ded79756ac744ac259f7d0df45da743e9077265","aliases":[],"snapshot":{"statement":"A nightly reprocessing job was accidentally deployed with this write and, on its first run, replaced an entire silver table with only a partial re-run's worth of data:\n\n```python\ndf.write.mode(\"overwrite\").saveAsTable(\"catalog.sales.silver_payments\")\n```\n\nThe team now wants every reprocessing job to fail loudly instead of touching a table that already exists. Which save mode gives that behavior?","options":["append, which replaces the table's previous contents with each new batch that gets written","overwrite, which refuses to write whenever the target table already contains any data at all","No save mode is needed, because Delta treats the modes as advisory hints and always appends the incoming data atomically","errorIfExists, the default save mode, which makes the write fail if the target table already exists"],"answer":"D","reference":"https://docs.databricks.com/aws/en/delta/best-practices","referenceTitle":"Best practices: Delta Lake — Databricks Docs"}}
{"id":"RAN-DBDEA-00233","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:293fe6c8d6a08f1f6a952799e7cc33be5f62f41dad2143f52e043c437cb7240b","aliases":[],"snapshot":{"statement":"A late-arrival job rebuilds only yesterday's slice of the events table, which is partitioned by event_date. In a single write, it must atomically replace only the rows whose event_date equals '2026-07-02' and leave every other day of the table untouched; the job's output DataFrame contains just the rebuilt day.\n\nWhich write approach meets both requirements?","options":["Run DELETE on the day's rows and then INSERT the corrected batch afterwards — each statement is atomic on its own, and together they cover the day","Write with mode('overwrite') and no further options — the whole table is replaced atomically, which also refreshes yesterday's rows","Write with mode('overwrite') and option('replaceWhere', \"event_date = '2026-07-02'\") — only the matching rows are replaced, in one atomic commit","Convert the table to external storage first, because replaceWhere cannot operate on Unity Catalog managed tables"],"answer":"C","reference":"https://docs.databricks.com/aws/en/delta/best-practices","referenceTitle":"Best practices: Delta Lake — Databricks Docs"}}
{"id":"RAN-DBDEA-00234","exam":"DBDEA","topics":["transformation"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:e1c3ef66008a9e146338071a6137cecb53b9c4e1c4b9b61782fce2cd901d1d74","aliases":[],"snapshot":{"statement":"Every night, a MERGE synchronizes catalog.sales.silver_orders with the source extract, updating matched orders and inserting new ones:\n\n```sql\nMERGE INTO catalog.sales.silver_orders AS t\nUSING catalog.sales.source_orders AS s\nON t.order_id = s.order_id\nWHEN MATCHED THEN UPDATE SET *\nWHEN NOT MATCHED THEN INSERT *\n```\n\nOrders are sometimes deleted in the source, and those deletions must also disappear from silver_orders, in the same atomic statement. Which clause adds this behavior?","options":["WHEN NOT MATCHED BY SOURCE THEN DELETE — it applies to the target rows whose keys have no counterpart in the source table","WHEN NOT MATCHED THEN DELETE — that clause already covers the target rows that are absent from the source","None: every MERGE clause can only act on rows of the source, so unmatched target rows are out of the statement's reach","Deletion is outside MERGE's scope, so schedule a separate DELETE ... WHERE order_id NOT IN (SELECT ...) after every merge runs"],"answer":"A","reference":"https://docs.databricks.com/aws/en/delta/merge","referenceTitle":"Merge data (Delta Lake upsert) — Databricks Docs"}}
{"id":"RAN-DBDEA-00235","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6c793b8d6a3bc6cec44cadcd7d983fb1a306ead33f9433f2693d0b08b9d75b6c","aliases":[],"snapshot":{"statement":"An ingestion job appends a DataFrame to catalog.sales.customers, an existing Delta table. This week the source schema added a new column, loyalty_tier, and the append now fails Delta's schema enforcement. The team wants the write to succeed by adding the new column to the table, and no existing rows should be rewritten.\n\nWhich write option achieves this?","options":["Use .option('overwriteSchema', 'true'), which adds the new column to the table while leaving every previously written row untouched","Use .option('mergeSchema', 'true'), which evolves the table schema additively — the new column is added with no rewrite","Use a TBLPROPERTIES change that turns schema enforcement off permanently, so that no write option is ever needed again","Use mergeSchema, keeping in mind that it also drops table columns that the incoming DataFrame no longer carries"],"answer":"B","reference":"https://docs.databricks.com/aws/en/delta/best-practices","referenceTitle":"Best practices: Delta Lake — Databricks Docs"}}
{"id":"RAN-DBDEA-00236","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:312f4aed4c68ee0112d6559681b0b73e5f1bfef1b2b6a8cb091a3ccecc6f0ebe","aliases":[],"snapshot":{"statement":"Years of analytics history live as Parquet datasets in the company's cloud storage lake. Governance now requires these datasets to become Delta tables, so that the lakehouse gains ACID transactions and time travel on top of them, but the data files must stay exactly where they are — no copies, no rewrites, no new storage locations.\n\nWhich approach meets the requirement?","options":["Run CREATE TABLE AS SELECT over each dataset, copying the rows into newly written Delta files under a managed location","Run CONVERT TO DELTA, which copies the existing Parquet files into the table's managed storage location first","Register the datasets as plain external Parquet tables instead — Parquet already guarantees ACID transactions, so converting them adds nothing","Run CONVERT TO DELTA on each dataset — Delta builds its transaction log over the existing files in place, without copying them"],"answer":"D","reference":"https://docs.databricks.com/aws/en/delta/","referenceTitle":"Delta Lake guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00237","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c04c2723fed67bcb23550cb74dfea71fc7e457783dfd265b3f87888c25a0977e","aliases":[],"snapshot":{"statement":"The development team wants a safe sandbox for risky experiments: a copy of the 100 TB production table catalog.sales.orders that they can transform, rewrite, and even corrupt without consequences. Storage costs must not double, and any commits made on the copy must never affect the production table.\n\nWhich approach fits?","options":["CREATE TABLE dev.orders AS SHALLOW CLONE catalog.sales.orders — the clone reuses the existing data files while keeping independent metadata and history","CREATE TABLE dev.orders AS DEEP CLONE catalog.sales.orders — a deep clone shares the underlying data files with the source table","CREATE TABLE dev.orders AS SELECT * FROM catalog.sales.orders — a full physical copy is unavoidable whenever dev writes must remain isolated from production","A shallow clone works only for read-only experiments: it keeps inheriting every later commit made on the source, so dev writes would flow back"],"answer":"A","reference":"https://docs.databricks.com/aws/en/delta/","referenceTitle":"Delta Lake guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00238","exam":"DBDEA","topics":["transformation"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9b5fc3b5043f2a86f35b816ceabc0876bcdf6fad468ac33bf35a19773762645f","aliases":[],"snapshot":{"statement":"A monthly reconciliation must compare the bank statement feed with the ledger: every row from either side must appear in the output, matched rows side by side, and rows present on only one side kept with NULLs on the other.\n\nWhich join produces this result?","options":["A FULL OUTER JOIN keeps rows from both sides, matching on keys and filling NULLs otherwise","An INNER JOIN on transaction_id, which guarantees that every row of both sides appears in the result","A LEFT JOIN of bank against ledger, which keeps unmatched ledger rows as NULLs","A UNION of both feeds, which places corresponding rows side by side for comparison"],"answer":"A","reference":"https://docs.databricks.com/aws/en/pyspark/reference","referenceTitle":"PySpark API reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00239","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:dc215bb17eba6242fc90aecc35e2e6b5b8c4771bd1f87685109ce4d727fdc73d","aliases":[],"snapshot":{"statement":"A Lakeflow Job starts with a notebook task, `pick_tables`, which scans operational metadata and computes the exact list of bronze tables that need a rebuild — a different list every night:\n\n```python\ndbutils.jobs.taskValues.set(key = \"tables\", value = tables)\n```\n\nThe next step must run the rebuild notebook once for every table in that list. How should the For each task that follows `pick_tables` be configured?","options":["Point its Inputs field at the task value reference {{tasks.pick_tables.values.tables}}, so the loop follows the list pick_tables produced","Copy the table names from last night's run into the Inputs field as a static JSON array, and edit the job whenever the list changes","Hardcode a master table list in the job definition, because For each inputs cannot reference values from other tasks","Drop the For each task and use a single notebook task that reads the array from pick_tables and loops over the tables with a Python while loop"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/tasks/for-each","referenceTitle":"Use a For each task to run another task in a loop — Databricks Docs"}}
{"id":"RAN-DBDEA-00240","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:2ed180061930ef7580cbd713fbb549e5ddafa1f19305ed712c8a290698765a40","aliases":[],"snapshot":{"statement":"An operations dashboard is published in the workspace, and the first analyst to open it each morning waits while all of its queries run from scratch. The team already runs a scheduled Lakeflow Job that finishes refreshing the underlying tables by 07:00, and they want the dashboard's results ready before analysts arrive at 08:00.\n\nWhat should the team add to this job?","options":["Nothing, because a published dashboard refreshes its results only when a user opens it, and no job task can change that","A Dashboard task after the table refresh, which refreshes the published dashboard's results as a step of the job","A notebook task that exports the dashboard to a PDF every morning, because only an export forces the dashboard's queries to run","A Dashboard task that publishes a new copy of the dashboard on every run, so analysts always open a freshly published version"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/configure-task","referenceTitle":"Configure and edit tasks in Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00241","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:fc9e3789178f8f89f999af47d4a532116209756528bd90fbb2c947f38c613a62","aliases":[],"snapshot":{"statement":"A Lakeflow Job runs two slow loads in parallel after the same extract task: load_customers and load_orders have no dependency on each other, and both must stay concurrent to fit the nightly window. A final task, run_quality_checks, must execute exactly once, only after both loads have completed successfully.\n\nHow should run_quality_checks be wired?","options":["Chain extract, load_customers, load_orders, and run_quality_checks in a strict line, because only linear chains can make a task wait for two others","Attach run_quality_checks to whichever load finishes first, and let it start as soon as that single branch completes","Split the checks into two tasks, one per branch, because a single task cannot depend on more than one upstream task","Make run_quality_checks depend on both load_customers and load_orders, keeping the default All succeeded run-if condition"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/configure-task","referenceTitle":"Configure and edit tasks in Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00242","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6f35d2a818acf609afd78fcdef7e01a98469b79997319a67932ce1763eb6b98e","aliases":[],"snapshot":{"statement":"A Lakeflow Job has six tasks — notebooks, one Python script, and one SQL file — that all live in the same GitHub repository on the release branch. Setting the repository and branch on every task is repetitive, and last month two tasks were accidentally pointed at different branches.\n\nHow should the job's source be configured?","options":["Configure the repository and branch once as the job's Git source, and have tasks reference only paths relative to the repository root","Repeat the full repository URL and branch on every task, since each task must carry its own complete source settings","Skip Git for this job, because remote Git sources can only be configured on individual tasks, never at the job level","Rewrite the SQL file task as a notebook task, because jobs that use a Git source cannot include SQL tasks"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00243","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:adc195f3ed193e0bfbe6f1f289fe107c60e9c5e05d9b54f69323bd958a2c5d90","aliases":[],"snapshot":{"statement":"A data engineer is editing a production Lakeflow Job in which the notebook task B already depends on the notebook task D. To make the two steps verify each other's output, the engineer adds a dependency in the opposite direction, configuring task D to depend on task B as well, and clicks Save.\n\nWhat is the result?","options":["The cycle is accepted, and the two tasks re-run in a loop until both finally succeed","The cycle is accepted, but the loop executes only once per run, after which the job continues with the downstream tasks","The dependency is accepted because both tasks are notebooks, and notebook tasks are allowed to close dependency cycles","The configuration is rejected: task dependencies must form a directed acyclic graph, and a cycle cannot be saved"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/configure-task","referenceTitle":"Configure and edit tasks in Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00244","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:dce4811f8144f94bb0742f96d43db207d9531ef19ae4b3f08eb35a21ef0f02e4","aliases":[],"snapshot":{"statement":"A data engineering team runs the same preparation pipeline for a morning report. Business users need it at 06:00 from Monday through Friday, while the Saturday data cut happens later, so the pipeline must run at 08:00 on Saturdays. The team wants one job definition, one run history, and one place to change the code.\n\nHow should the job be configured to fire on both schedules?","options":["Add two scheduled triggers to the job: a weekday 06:00 cron and a Saturday 08:00 cron","Clone the job into two identical jobs, each with its own schedule, and maintain both in parallel","Keep a single trigger, because one Quartz cron expression can fire at different times on different days","Leave the weekly schedule as-is and rely on its automatic retries to produce the Saturday 08:00 run"],"answer":"A","reference":"https://docs.databricks.com/aws/en/jobs/triggers","referenceTitle":"Automate jobs with schedules and triggers — Databricks Docs"}}
{"id":"RAN-DBDEA-00245","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:aa2cd1b15d00ae621152e277b0fe9731c2324d9e238e5b45e6c11595e5a53ef3","aliases":[],"snapshot":{"statement":"A platform team's master Lakeflow Job coordinates the company's onboarding flow. One step must launch the partner team's onboarding job, wait for its outcome, and let downstream tasks in the master job run only if that job succeeded. The partner's job stays owned and scheduled by the partner team.\n\nHow should the master job trigger the partner's job?","options":["Use a notebook task that calls the Jobs REST API to start the partner's job and poll for its result","Use a pipeline task pointing at the partner's job, since pipeline tasks can execute any existing job","Use a Run Job task that starts the partner's job, so its outcome becomes a step in the master DAG","Use a manual handoff, because Lakeflow Jobs cannot launch other jobs in a managed way"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/configure-task","referenceTitle":"Configure and edit tasks in Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00246","exam":"DBDEA","topics":["jobs"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:62113eee28842858d072f560d47d65a76a3902245d91d86674116601b1ab799e","aliases":[],"snapshot":{"statement":"A Lakeflow Job declares a job-level parameter target_date, and one of its notebook tasks declares a task parameter with the same key target_date but a different default. An on-call engineer then triggers the job with Run now, passing their own value for target_date.\n\nWhich value does that task receive?","options":["The task parameter's default, because task-level settings are always the most specific","The job parameter's default, because job-level settings override everything including Run now","An error at run time: declaring the same parameter key at two levels is rejected by the platform","The Run now value wins; below that, the job-level parameter beats the task parameter on a shared key"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00247","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:90a190c12f237c4df84dd585da22d861716fc27b285ab9b69e1624c63dab89c3","aliases":[],"snapshot":{"statement":"Governance rules at a company state that compute used by Lakeflow Jobs must follow approved shapes: permitted node types and a maximum worker count. The platform team wants these rules enforced automatically whenever engineers configure compute for their jobs, not reviewed case by case.\n\nWhat should the platform team do?","options":["Apply a compute policy to the job compute, so permitted node types and size limits are enforced","Document the approved shapes in the team wiki, because compute policies apply only to interactive all-purpose compute","Rely on automatic inheritance, because every job takes the compute policy of the engineer who created it","Cap the number of clusters per workspace, because compute policies control only how many clusters can exist"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00248","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:44585ee919402b7a245280cd22651e6aeb1a4c993b42248b67924d25bc5ce7f3","aliases":[],"snapshot":{"statement":"A team is migrating an old ETL process to Lakeflow Jobs. Today one long notebook chains five processing stages with %run: each stage calls the next notebook, everything shares a single execution, and a failure in stage three means rerunning the whole chain the next morning.\n\nWhat is the main benefit of rebuilding this chain as separate notebook tasks in one job?","options":["The ability to keep a single shared execution, since chained %run notebooks are the recommended pattern for production","The ability to keep the %run chain, because jobs cannot run notebooks as separate tasks","The ability to retry, monitor, and branch each stage independently, and rerun only the failed stage","The ability to preserve %run behavior exactly, with less configuration to maintain than separate tasks"],"answer":"C","reference":"https://docs.databricks.com/aws/en/jobs/tasks/notebook","referenceTitle":"Notebook task for jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00249","exam":"DBDEA","topics":["jobs"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:deb32f3f4249208fa79d0585cb5947eaa589d8b558d0c01fe79695260619885f","aliases":[],"snapshot":{"statement":"A source system that feeds a nightly Lakeflow Job will undergo maintenance tomorrow, and the team decides the job should not fire while the data is unstable. Today's run has already started and is still in progress. The job definition, its run history, and next week's schedule must all be preserved.\n\nWhat should the team do?","options":["Cancel the in-progress run immediately, because pausing a job also kills every run that is still active","Pause the job's scheduled trigger: it stops starting new runs, and the active run still finishes","Remove the schedule from the job definition and re-enter it by hand after the maintenance window","Delete the job and rebuild it after maintenance, because stopping firings requires deleting the job"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/configure-job","referenceTitle":"Configure and edit Lakeflow Jobs — Databricks Docs"}}
{"id":"RAN-DBDEA-00250","exam":"DBDEA","topics":["jobs"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:6fb33eec6deacdc5e88e05e5229acfe9f4c4c51cb87ceb911bab0dbf38c3074f","aliases":[],"snapshot":{"statement":"The first task in a Lakeflow Job runs a SQL task that computes yesterday's totals; a downstream notebook task needs that number as a parameter, but the SQL task does not use dbutils, so it cannot set a task value itself.\n\nHow can the SQL task's output reach the notebook task?","options":["Have the SQL task write the value with dbutils.jobs.taskValues.set in a trailing statement","Pull it with a dynamic value reference like {{tasks.totals.output.first_row.yesterday_total}}","Recompute the total in the notebook with an identical SQL query against the same tables","Declare the number as a job parameter and edit the job definition manually every single night"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/dynamic-value-references","referenceTitle":"Use dynamic value references — Databricks Docs"}}
{"id":"RAN-DBDEA-00251","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f1c404a4ad1b6753354c15f3216a5f41471fe2ffd4d52b975d7a997ddb3d63dd","aliases":[],"snapshot":{"statement":"A downstream Lakeflow Job must refresh whenever two specific upstream tables change — sales.orders and sales.returns — but a third table in the same schema, sales.staging, changes constantly and must never trigger the job.\n\nHow should the table update trigger be configured?","options":["Monitor the whole sales schema, and cancel any run that was triggered by sales.staging","Select sales.orders and sales.returns as the trigger's monitored tables, so only they fire the job","Add a filter in the job's first task that checks which table changed before continuing","Rely on the default behavior, because table update triggers always watch every table in the source schema"],"answer":"B","reference":"https://docs.databricks.com/aws/en/jobs/trigger-table-update","referenceTitle":"Trigger a job when a table updates — Databricks Docs"}}
{"id":"RAN-DBDEA-00252","exam":"DBDEA","topics":["jobs"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:bf47f9460f04303328d78a0145d4749b6887e285ce3710f38ccdce7194a582ca","aliases":[],"snapshot":{"statement":"A team configures a file arrival trigger by pasting the raw cloud URL of an S3 bucket where a vendor drops CSV files: s3://vendor-drops/incoming/. The trigger setup fails, and the team asks what location it can actually monitor.\n\nWhat is the requirement?","options":["Any s3:// path works as long as the bucket is in the same region as the workspace","Triggers can only watch workspace DBFS paths, so the vendor must upload everything to /FileStore instead","The raw bucket URL works if the team also pastes AWS access keys into the job definition","File arrival triggers monitor a Unity Catalog volume or registered external location path, not arbitrary URLs"],"answer":"D","reference":"https://docs.databricks.com/aws/en/jobs/file-arrival-triggers","referenceTitle":"Trigger a job when a file arrives — Databricks Docs"}}
{"id":"RAN-DBDEA-00253","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7c54f19e35424d4db2f08d422057bd5e5b296eb1844bb650d5f5d23457311195","aliases":[],"snapshot":{"statement":"A data engineer fixes a null-handling bug in the silver_orders notebook inside the team's Git folder, which is cloned from the team's remote repository and has the feature/null-fix branch checked out. They finish the edit and must send the change to the remote repository so a teammate can review it, without leaving the Databricks workspace. What should the engineer do?","options":["Open the Git dialog, review the highlighted changes, add a commit message, and click Commit & Push to send the branch to the remote","Push the change with Git on a local machine, because the workspace UI can edit notebooks but cannot push commits","Save the notebook and wait: the Git folder commits every change automatically and pushes it to the checked-out branch","Use the Git dialog only to browse the branch history, because it shows past commits but cannot create new ones"],"answer":"A","reference":"https://docs.databricks.com/aws/en/repos/git-operations-with-repos","referenceTitle":"Git operations with Git folders — Databricks Docs"}}
{"id":"RAN-DBDEA-00254","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:76f05723f89dff0888d30efe48848617580ac32228b2f1b7c28ee618ea540b39","aliases":[],"snapshot":{"statement":"A data engineer uses the Databricks CLI on a laptop that runs commands against both the team's development workspace and the production workspace. They are tired of re-typing credentials and want a fast, explicit way to choose which workspace a command targets. How should the engineer configure the CLI?","options":["Rely on the browser session signed in to the workspace, since the CLI can only reach that one workspace","Add a named configuration profile per workspace to ~/.databrickscfg and select it with --profile or -p","Accept a single host per machine, because the CLI supports only one workspace URL in its configuration","Avoid environment variables entirely, because DATABRICKS_HOST and DATABRICKS_TOKEN are not recognized by the CLI"],"answer":"B","reference":"https://docs.databricks.com/aws/en/dev-tools/cli","referenceTitle":"Databricks CLI — Databricks Docs"}}
{"id":"RAN-DBDEA-00255","exam":"DBDEA","topics":["cicd"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:2e67bcc1a0755737c1010724e4dc2743af7a32bf6daad5d24e16a9f7c92fc900","aliases":[],"snapshot":{"statement":"A team's CI log shows this command completing successfully, with no target specified:\n\n```bash\ndatabricks bundle validate\n```\n\nA junior engineer expected the command to fail without -t and is surprised that the output mentions target dev. Why did the command validate the dev target?","options":["It validates all targets declared in the bundle configuration one after another in a single pass","It errors out, because every bundle command requires an explicit --target option","It uses the target marked default: true in the bundle configuration, which is dev here","It always validates the production target first, because that is the safest default"],"answer":"C","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00256","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9870d3a7af6c28ab5ea2cd6424aae1c1b28c1227f5404d3758ae015dc2c40608","aliases":[],"snapshot":{"statement":"A team's Automation Bundle defines one Lakeflow Job and its notebooks. The team promotes releases through development, staging, and production workspaces and must guarantee that the exact code revision that passed staging tests is what runs in production. Which promotion approach provides that guarantee?","options":["Keep a separate branch per environment, so each workspace always deploys its own copy of the code","Merge the development branch straight into the production branch whenever a release is due","Deploy staging's workspace files to production by copying the tested folders in the workspace UI","Deploy the same bundle commit to each target in turn; only target-level variables and overrides differ between environments"],"answer":"D","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00257","exam":"DBDEA","topics":["cicd"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6698cea45124e76a95f773334bf5f55fd032e77da5073ae2b0f9546f676b0b5e","aliases":[],"snapshot":{"statement":"A company's data platform runs on Bitbucket, and the platform lead wants to manage pipeline notebooks in Databricks Git folders connected to the company's Bitbucket repositories. Before planning the rollout, the lead asks which Git providers Git folders support. What is the correct answer?","options":["GitHub, GitLab, Azure DevOps (Azure Repos), Bitbucket Cloud, and Bitbucket Server and Data Center are all supported","Only GitHub repositories can be connected, so the team must mirror its Bitbucket repositories to GitHub","GitLab and Bitbucket require a third-party connector purchased from a Databricks partner","Any Git hosting service works with no restrictions, including self-hosted servers on any network"],"answer":"A","reference":"https://docs.databricks.com/aws/en/repos/git-folders-concepts","referenceTitle":"Git folders concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00258","exam":"DBDEA","topics":["cicd"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:57d42eefbff434edd535963f90e2bf4e415f6c624ded7988bbb3bd960ad515b2","aliases":[],"snapshot":{"statement":"A team deploys one Automation Bundle to three targets: dev, test, and prod. Each target is a different Databricks workspace, and the team must keep each environment's deployed files and state separate so deployments never collide. Where should the per-environment workspace host and paths be defined?","options":["In fixed global paths that every target shares, because one bundle uses one workspace location","In three separate bundles, one per environment, each shipping its own folder layout","In each target's workspace mapping, declaring that environment's host together with root_path and artifact_path","Nowhere: the workspace is chosen by whichever configuration profile happens to be active in the deployer's terminal"],"answer":"C","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00259","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:4874b5c4e59972d207626fed60e6d4ee522e811576b26769003663f387cd403a","aliases":[],"snapshot":{"statement":"An Automation Bundle deploys a Lakeflow Job whose Python task must call an external API with a bearer token. Security policy forbids the token from appearing in any versioned file or in the deployed job's visible configuration. The bundle files live in Git. How should the task get the token?","options":["Declare a bundle variable for the token and pass its value with --var each time the pipeline deploys","Point a task environment variable at a secret scope with a reference like {{secrets/scope/key}}, resolved at run time","Commit a placeholder string in the YAML and have the CI pipeline rewrite the file with sed right before the deploy step runs","Keep the token directly in the YAML, since the team's Git repository is private to the platform group"],"answer":"B","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00260","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ffe72e44b345533a1b1365fc8cba1de211ea02d7f338bd85748fc409d12ae60d","aliases":[],"snapshot":{"statement":"A company keeps all of its data products in one large monorepo. The orders team only needs the /products/orders directory — its notebooks and job definitions — available in a Databricks Git folder, not the entire repository. How should the team create the Git folder?","options":["Create it normally: a Git folder always clones the full repository, and unused directories must simply be ignored","Tell the team that monorepos are not supported and the product must move to its own repository","Maintain a mirror repository that contains only the orders product, and connect the Git folder to the mirror","Enable sparse checkout when creating it and set the cone pattern to /products/orders, cloning only that subset"],"answer":"D","reference":"https://docs.databricks.com/aws/en/repos/git-folders-concepts","referenceTitle":"Git folders concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00261","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e75c3ce7bfbe0a4e10803909dfc2f4f9105ae52e8a6921b6096ad337d1dd14cb","aliases":[],"snapshot":{"statement":"While developing ingestion notebooks locally in an IDE, a data engineer wants their local project folder and a folder in the Databricks workspace to stay in step as files change, and asks whether the CLI's sync command keeps both sides identical. What does databricks sync actually do?","options":["It synchronizes one way, local directory to workspace folder; edits made in the workspace never flow back","It synchronizes in both directions, keeping the local project folder and the workspace folder identical at all times","It replaces Automation Bundle deployments, pushing project files straight into production workspaces","It copies local files into DBFS paths such as dbfs:/tmp, since it cannot write to workspace folders"],"answer":"A","reference":"https://docs.databricks.com/aws/en/dev-tools/cli","referenceTitle":"Databricks CLI — Databricks Docs"}}
{"id":"RAN-DBDEA-00262","exam":"DBDEA","topics":["cicd"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:f7b1b59d3e0bb503066eaf470c16806beeea0e2400d9821596c1d14a1bfb34fa","aliases":[],"snapshot":{"statement":"A new data engineer has only seen the Databricks CLI used to deploy Automation Bundles and assumes that bundles are all it is for. The team lead wants to correct this with an accurate description of what the CLI covers. Which description is accurate?","options":["The CLI is a bundle-only tool: it validates, deploys, and runs Automation Bundles and nothing else","The CLI runs exclusively inside the Databricks workspace and cannot be installed on a laptop","The CLI wraps the platform's REST APIs, managing clusters, jobs, pipelines, files, permissions, and more","The workspace UI can do everything the CLI does just as conveniently, so the CLI adds no real capability for teams"],"answer":"C","reference":"https://docs.databricks.com/aws/en/dev-tools/cli","referenceTitle":"Databricks CLI — Databricks Docs"}}
{"id":"RAN-DBDEA-00263","exam":"DBDEA","topics":["cicd"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:905c765fbbe762e9b22bedb6890281f20ee0c3ccd596b3de400ffb9a6c968070","aliases":[],"snapshot":{"statement":"A team's Automation Bundle defines a Lakeflow Job whose notebook task references a project file, notebook_path: ../src/ingest.py. The bundle deploys to the dev target for the first time, and an engineer asks where ingest.py ends up so they can trace what the job actually runs. Where do the bundle's files go at deploy time?","options":["The job reads the file straight from the Git provider at run time, so no copy ever lands in the workspace","The deploy uploads the files into the target workspace's bundle paths, and the deployed job points at those copies","The files stay on the engineer's machine, and the job fetches them from there whenever it runs","The deploy creates a Git folder in the target workspace that holds the project's files for the deployed job to pull from"],"answer":"B","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00264","exam":"DBDEA","topics":["cicd"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:fcddab940ea27dae27d90ce5abe99de446b96e3eb2323a36ee859a0fef2a8ace","aliases":[],"snapshot":{"statement":"Minutes after a routine deploy of the prod target, dashboards show that a change in the release broke a downstream metric. The team must restore production to the last-known-good version of the bundle quickly and cleanly. What should they do?","options":["Run databricks bundle rollback -t prod to restore the previous deployment automatically","Delete the affected jobs and pipelines in the production workspace and let the owning teams rebuild them by hand","Select the last-known-good deployment from the versions the bundle keeps in the workspace and re-apply it","Revert the bad commit in Git and redeploy prod; the bundle then overwrites the resources with the restored code"],"answer":"D","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00265","exam":"DBDEA","topics":["cicd"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:0e0d4382c3080f285aa5e54354fb5d3a94a4350dd96146208f7d3ae3cde30e2e","aliases":[],"snapshot":{"statement":"One bundle serves several CI pipelines, and each runner needs a slightly different value for the same setting — for example, the catalog prefix the deployed pipeline writes to. The team wants to keep a single databricks.yml and avoid per-pipeline copies of the configuration. How can the differing values be supplied?","options":["Declare the differing values as bundle variables and supply each runner's values through BUNDLE_VAR_-prefixed environment variables","Hardcode every value, because bundle configuration files cannot read anything from the environment","Generate a different databricks.yml for each pipeline, with each runner's value written into its own file","Use environment variables only to choose the workspace host, because that is the only setting they can affect"],"answer":"A","reference":"https://docs.databricks.com/aws/en/dev-tools/bundles/","referenceTitle":"What are Databricks Automation Bundles — Databricks Docs"}}
{"id":"RAN-DBDEA-00266","exam":"DBDEA","topics":["cicd"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:d6df2236a9a9442a2b63e7bb3b4077b436360d2eddf44dd65045fb5e27ecd5cb","aliases":[],"snapshot":{"statement":"A GitHub Actions runner runs Databricks CLI commands on every merge to main. DATABRICKS_HOST is already set to the workspace URL, the team has created the service principal ci-release with an OAuth secret, and the workflow must act as that service principal with no interactive sign-in. What must still be configured on the runner?","options":["A weekly calendar reminder for an engineer to run databricks auth login interactively on the runner","The username and password of a platform engineer as workflow variables for the CLI to authenticate with","The service principal's client ID and secret as DATABRICKS_CLIENT_ID and DATABRICKS_CLIENT_SECRET, so the CLI runs OAuth M2M","A personal access token that belongs to a human administrator, because service principals cannot authenticate the Databricks CLI"],"answer":"C","reference":"https://docs.databricks.com/aws/en/dev-tools/cli","referenceTitle":"Databricks CLI — Databricks Docs"}}
{"id":"RAN-DBDEA-00267","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:45944fcc41319e2dd360687f6215af330763b8c500ed48cddc83a82510e1f48c","aliases":[],"snapshot":{"statement":"A data engineer is new to the Databricks platform and wants to study a slow PySpark notebook run in the Spark UI. The engineer wants to start from the actions the notebook triggered, open the slowest one, and then see how many tasks each of its stages ran and how long every task took. Which navigation matches how the Spark UI organizes this information?","options":["The Jobs tab lists the triggered actions; a job page shows its stages; the Stages view lists each stage's tasks with durations and I/O metrics","The Storage tab reports how long every task ran inside each stage, next to the cached tables that those tasks read","The Executors tab displays the execution plan of each action, including stage boundaries and per-task assignments","The Jobs and Stages tabs are the same view, with Stages only adding extra filters by duration and status"],"answer":"A","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/","referenceTitle":"Spark UI guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00268","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:0d4e5bd6df3778ede93dc7332fa303433ace99239c106fdf23b447dd3fa4b885","aliases":[],"snapshot":{"statement":"An all-purpose cluster that hosts the team's interactive notebooks was restarted overnight: sessions died, and the uptime counter had reset by morning. The engineer did not schedule any restart and wants to reconstruct exactly what happened to the compute while nobody was watching — whether it was resized, terminated, or failed, and at what time. Where is that record kept?","options":["In the job run history, which records every cluster lifecycle event — creations, resizes, terminations — of the compute that jobs used","In the cluster event log, which is written only for job clusters and never for all-purpose compute","In the cluster event log, which records timestamped events such as creation, resizes, terminations, and failures","In no view at all, because the platform does not persist cluster-level events like restarts or resizes"],"answer":"C","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00269","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:dc39d63e42c84d6bae6319fd3f7a30284f57c0adad32d164018bd495c463e8ae","aliases":[],"snapshot":{"statement":"Databricks SQL dashboards have been intermittently slow for a week. The team lead wants to review the slowest queries that one specific analyst ran on the shared warehouse last Tuesday, see how long each took, and check whether they were queued or spilling. How can the lead find them without leaving the workspace UI?","options":["Ask the analyst to export her own list manually, since each user's history is private and no consolidated view exists for the team","Open Query History, filter by user, warehouse, and date range, then sort the runs by duration to surface the slowest","Open a query profile, which lists every query in the workspace ranked by elapsed time and spill","Rerun each dashboard query by hand, because query history is discarded as soon as execution finishes"],"answer":"B","reference":"https://docs.databricks.com/aws/en/sql/user/queries/query-history","referenceTitle":"Query history — Databricks Docs"}}
{"id":"RAN-DBDEA-00270","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:04cce8c74337c5ff7174ae8e33ca83c502dc20aa8708c26e95fc0cdd52100564","aliases":[],"snapshot":{"statement":"A dashboard query against two gold Delta tables finished in about 40 seconds when the warehouse was cold. Minutes later, the exact same query returned in two seconds, although no files had changed. The team concluded that the warehouse 'warmed up' and now expects every rerun to be fast. What actually explains the second run?","options":["The result cache is global to the account and never expires, so any user running any similar query gets the two-second answer","The cached result also covers queries with different filters over the same tables, which are answered without re-execution","The query was re-executed in the background at higher priority, and the cache only duplicates its profile in query history","The result cache served the identical query without re-execution, since the underlying tables had not changed"],"answer":"D","reference":"https://docs.databricks.com/aws/en/sql/user/queries/query-profile","referenceTitle":"Query profile — Databricks Docs"}}
{"id":"RAN-DBDEA-00271","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c2431d1965daa812b1a3b773ad3f99d908060b594431a67ccdc923b50ebb7872","aliases":[],"snapshot":{"statement":"A streaming workload ingests Kafka events into a streaming table through a continuously running query. Throughput seemed to drop after a configuration change, and the engineer wants to confirm, on the live query, how many input rows per second the source is delivering and how long each micro-batch takes to process. Where should the engineer look?","options":["The Structured Streaming tab of the Spark UI, which shows active queries with input rate and batch duration statistics","No Spark UI tab, because streaming queries are deliberately excluded from every view of that interface","A StreamingQueryListener registered in the query's code, which is the only way to observe input rates and batch durations","The Lakeflow job run history page, the sole view that reports per-query stream metrics like rows per second"],"answer":"A","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/","referenceTitle":"Spark UI guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00272","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:cf9b62114e64559ab0370de19fa5b2b3e5b0fbde72f05786da7d6196effb35a4","aliases":[],"snapshot":{"statement":"A job joins a large events table with a small users table on `account_id`, and a handful of accounts hold most of the rows. The Spark UI shows one straggler task per run, and AQE skew handling never triggers because the oversized partition stays below both size thresholds. Which technique redistributes the hot key manually?","options":["Increase `spark.sql.shuffle.partitions` far above the current value, so that the single hot key's rows spread across many smaller tasks","Broadcast the large events table to every executor, so each task joins its local slice against the users table","Append a random salt value (0 to N) to `account_id`, replicate each users row across every salt, and join on the composite key","Repartition the events table by `account_id` before the join, so all rows of each key are co-located in one partition"],"answer":"C","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/long-spark-stage","referenceTitle":"Investigate long-running stages in the Spark UI — Databricks Docs"}}
{"id":"RAN-DBDEA-00273","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:a00e5b1c31ebaa5af4f5d6c0769eb2f700a7485688d6e6d171c6a6dcb5dcbdc7","aliases":[],"snapshot":{"statement":"A nightly aggregation shuffles about 6 GB of data. The plan submitted to the cluster ended with 200 post-shuffle partitions, yet the Spark UI shows the final stage ran only 12 tasks, each processing a reasonably sized slice — and nobody edited the query. Which mechanism changed the task count while the query was running?","options":["The driver, which locks the final task count when the plan is compiled and submitted for execution","Adaptive Query Execution, which coalesces small post-shuffle partitions into fewer, larger tasks using runtime statistics","A manual repartition transformation, which engineers are required to add to every job before Spark can merge its small shuffle partitions","Photon, whose vectorized engine merges small partitions while it executes the stage's scans and aggregations"],"answer":"B","reference":"https://docs.databricks.com/aws/en/optimizations/aqe","referenceTitle":"Adaptive Query Execution — Databricks Docs"}}
{"id":"RAN-DBDEA-00274","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9c4f8299ca5d6b895890ff3161874ed58a7f234010730ecd8b983b8ff9a86e43","aliases":[],"snapshot":{"statement":"A PySpark job reads a large Delta table that was written as many small files. The read stage schedules over 4,000 tasks that each finish in a fraction of a second, and most of the stage time goes to scheduling overhead. Which setting controls how much data each read task processes, so the small files can be packed into fewer tasks?","options":["`spark.sql.files.maxPartitionBytes`, which however only influences the size of the files written by the job's outputs","The cluster's core count, because input splits are derived solely from the number of worker nodes and no tunable target size exists","The split target, which must never be raised, because larger tasks always degrade reads in every workload","`spark.sql.files.maxPartitionBytes`, which sets how many bytes are packed into a single read partition (128 MB by default)"],"answer":"D","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/","referenceTitle":"Spark UI guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00275","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:bc39e2e4c484d563a070c7f15536b2a410d63864153f3f8614e6dc4d4334fd50","aliases":[],"snapshot":{"statement":"A sales fact table is partitioned by `order_date`. A query filters `WHERE order_date = '2026-10-02'`, but it still scans hundreds of gigabytes and runs for minutes. Before touching the table, the engineer wants to confirm whether the partition filter was actually applied to the scan. How can the engineer verify this in the query plan?","options":["Run the query and inspect the Ganglia charts afterward, the only place where partition pruning statistics are exposed","Accept that pruning is an internal optimizer step that never becomes visible in any plan output","Run EXPLAIN, or open the plan in the query profile, and check the scan node for the partition predicates applied","Skip the check, because partition pruning applies automatically to filters on any column of the table"],"answer":"C","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/","referenceTitle":"Spark UI guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00276","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:35c1847510a39162c913754efb8ba80b0fb44093402996e20099109c690c5ed3","aliases":[],"snapshot":{"statement":"A data engineer supports dozens of Databricks SQL users and keeps finding the same performance problems: tables fragmented into small files, exploding joins, and skewed aggregations. The engineer wants a place where the platform itself flags these issues for specific queries and suggests what to do about them. Where should the engineer look?","options":["Performance insights, which appear in query history and the query profile with actionable recommendations","Databricks Support tickets, because query recommendations are authored only by the support team on request","The insights page, which makes modeling decisions like schema and partitioning choices on the engineer's behalf","Cached query results, the only executions for which the platform generates performance insight recommendations"],"answer":"A","reference":"https://docs.databricks.com/aws/en/sql/user/queries/performance-insights","referenceTitle":"Query performance insights — Databricks Docs"}}
{"id":"RAN-DBDEA-00277","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:874dfeeed413514024211558ff8065f6e2035607634d47b2cdb5dd384ed20f05","aliases":[],"snapshot":{"statement":"In the jobs timeline, one stage of a nightly Lakeflow Job takes 40 minutes while its sibling stages finish in about three minutes each. Before changing any configuration, the engineer wants to understand why this particular stage is slow. What is the recommended next step in the investigation?","options":["Resize the cluster to a larger and more expensive node type right away, since a long stage always means the hardware is undersized","Open the stage page and examine the task duration distribution, summary metrics for skew, and spill statistics","Assume the notebook code has a bug, because stages only run long when user code contains an error","Cancel and resubmit the run, because re-execution is the documented way to clear a long-running stage"],"answer":"B","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/long-spark-stage","referenceTitle":"Investigate long-running stages in the Spark UI — Databricks Docs"}}
{"id":"RAN-DBDEA-00278","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:8183c6d8b4c12bd220cc2dacf172d9912a817ad03f8b8a31fa541bad5300593f","aliases":[],"snapshot":{"statement":"A complex batch ETL join stage spills over 40 GB of shuffle data to local disk on every nightly run. The query cannot be redesigned this quarter, so the team wants a compute-level change that reduces the spill without touching the code. Which change follows the documented guidance?","options":["Treat the spill as harmless log noise, because writing shuffle data to disk never slows a stage down","Reduce the memory per executor, which forces Spark to launch more tasks and split the shuffle data among them","Disable Adaptive Query Execution, which removes partition coalescing and with it all spill from the stage","Use instances with more memory, or fewer and larger workers, so more of the shuffle fits before spilling"],"answer":"D","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/","referenceTitle":"Spark UI guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00279","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:770c19abad369d57e4de964e9d79734a36586c36fc0c8d79159de39cb0edca31","aliases":[],"snapshot":{"statement":"A team changed one shuffle configuration on their nightly pipeline, and the on-call engineer reports that the job 'feels faster' now. Before rolling the change out to the other pipelines, the lead wants proof that the tuning actually improved the slow stage. What is the correct way to verify it?","options":["Run the same workload before and after the change, then compare stage duration and spill in the Spark UI","Change several more tuning parameters at once, then judge the combined effect on the pipeline's total runtime","Trust the on-call engineer's impression, because perceived runtime is a reliable signal of pipeline performance","Compare Git commit timestamps around the change, which serve as a proxy for how long each pipeline run took"],"answer":"A","reference":"https://docs.databricks.com/aws/en/optimizations/spark-ui-guide/","referenceTitle":"Spark UI guide — Databricks Docs"}}
{"id":"RAN-DBDEA-00280","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:6b7bc0fad6310235a0a4825183d0eb7892850dd1af6b7a0947617e9e3806973d","aliases":[],"snapshot":{"statement":"A pro SQL warehouse serves the company's morning dashboards. At 9 a.m., more concurrent queries arrive than the running clusters can execute, and analysts watch their queries sit in QUEUED status. The warehouse is configured with a minimum of 1 and a maximum of 6 clusters. What does the warehouse do while queries are queued?","options":["It keeps them queued and provisions additional clusters of the same size, up to the configured maximum, to drain the backlog","It replaces each cluster's driver with a larger driver type, which frees capacity for every waiting query","It fails the queued queries immediately with a timeout, forcing analysts to resubmit after the morning peak","It keeps growing past the configured maximum, because warehouse scale-out has no configurable limit"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute/sql-warehouse/warehouse-behavior","referenceTitle":"SQL warehouse sizing, scaling, and queuing behavior — Databricks Docs"}}
{"id":"RAN-DBDEA-00281","exam":"DBDEA","topics":["troubleshooting"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:19383e2d28190a53dd8119d739954ae4100d1ddcca1520c58963749310cc9070","aliases":[],"snapshot":{"statement":"A long-running job on a cluster with spot workers loses one mid-run: the node is reclaimed by the cloud provider, and a shuffle stage that had tasks on that node is suddenly incomplete. A new engineer assumes the whole job run is lost.\n\nWhat actually happens?","options":["Spark re-executes the lost tasks and dependent stages, so the run continues without starting over","The entire job run fails immediately, and Lakeflow Jobs restarts the whole thing from the first task","The lost tasks are skipped, so the run completes with silently missing partitions","Spot reclamation pauses the job until the same node type becomes available again"],"answer":"A","reference":"https://docs.databricks.com/aws/en/compute/configure","referenceTitle":"Configure clusters — Databricks Docs"}}
{"id":"RAN-DBDEA-00282","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:103879f886adec2b00ca97dd5bff17048ad69bb3082e050d150edba9667ae217","aliases":[],"snapshot":{"statement":"An external security auditor is reviewing how the company's lakehouse reaches its cloud storage. The platform lead explains that one Unity Catalog object encapsulates the cloud credential — an AWS IAM role — that authorizes access to storage paths, and the auditor asks who in the platform can create that object. Which explanation is accurate?","options":["Any lakehouse user can create a storage credential whenever they register a new cloud storage path for their team.","A storage credential is created at the metastore level by a metastore admin or a principal granted CREATE STORAGE CREDENTIAL.","Storage credentials live inside individual schemas, next to the tables and volumes that read from their storage paths.","Storage credentials are account-wide objects that every user shares automatically, with no Unity Catalog privilege required to use them."],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00283","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:425780c2d0c09829725a353551b0be66789ab1562d0eea0662328757e1105fd9","aliases":[],"snapshot":{"statement":"A new analytics domain is joining the lakehouse, and the governance team plans to give it a dedicated Unity Catalog catalog so its schemas and tables stay separate from other domains. The platform admin who will create the catalog asks which privilege the task requires. Which answer is correct?","options":["CREATE CATALOG on the metastore, a privilege that metastore and account admins can grant to other principals.","USE CATALOG on an existing catalog, since usage privileges on one catalog authorize creating sibling catalogs.","Workspace admin in every workspace attached to the metastore, because catalog creation is a workspace-scoped operation.","SELECT on at least one existing catalog, because demonstrated read access is the prerequisite for creating new catalogs."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00284","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ccc9824c3d0f6636273eacd569154e96fa2c71849d5072bd6acec162e25046c2","aliases":[],"snapshot":{"statement":"A data engineer created the Unity Catalog table main.finance.risk_scores and owned it until she changed teams. No GRANT statement ever named her user, yet she could always query the table, modify its data, and grant other groups access to it. Her successor asks how that was possible. What is the explanation?","options":["Object owners receive hidden grants that SHOW GRANTS deliberately omits from its output to keep listings short.","An owner's only automatic capability is dropping the table; reads and writes still require explicitly granted privileges.","Unity Catalog grants owners every capability on their object implicitly — including granting privileges to other principals.","Ownership cannot be transferred after creation, so the engineer's account must keep owning the table forever."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00285","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:17dc3ae2a2747142c6edaf5563c2bcfec33447f2084daa673e9fc74eff6fb10b","aliases":[],"snapshot":{"statement":"A governance program is rolling out governed tags — key-value classifications such as pii=high — so that ABAC policies can act on them uniformly. The data governance lead is documenting where these tags are defined and who can create them.\n\nWhich statement matches the platform?","options":["Governed tags are defined per metastore, and any metastore admin can create them freely","Governed tags are created ad hoc by users on their own tables, with no central control","Governed tags are account-level, applying across the whole estate including multiple metastores","Governed tags are defined per workspace and never shared between workspaces"],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/abac/","referenceTitle":"Attribute-based access control — Databricks Docs"}}
{"id":"RAN-DBDEA-00286","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3083760fbb6bac27b524754aa33b5343a3e2b9cdd2d80a0255b3b3fcc33c2bd2","aliases":[],"snapshot":{"statement":"The governance team plans to rename msrp, a single column in main.products.price_list, a 60-column table with dozens of downstream dashboards and pipelines. Most consumers never touch that column, so the team wants to reach only the assets that actually depend on it, instead of every consumer of the table. Which Unity Catalog capability supports this?","options":["Unity Catalog tracks lineage at the table level only, so every consumer of the table counts as affected.","Unity Catalog captures lineage down to the column level, and Catalog Explorer shows which assets depend on that column.","Column lineage becomes available only after an administrator enables audit logging on each table that needs tracking.","Column-level lineage is captured only for external tables, because their files live outside the managed storage location."],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00287","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:42646ffeee96dea76e2ca5e3a418dba9d834adffe091897e8aeeb1b8132f8b86","aliases":[],"snapshot":{"statement":"A platform team is designing file storage under Unity Catalog. Machine learning artifacts produced by Databricks jobs should live in storage that Unity Catalog fully controls, including deletion. A large reference dataset already sits in a bucket owned by another application, must stay there, and must never be deleted by Databricks. How should the team create the two volumes?","options":["Create both as external volumes, because Unity Catalog volumes must always reference paths governed by an external location.","Create both as managed volumes, each with an explicit LOCATION clause that names its destination storage path.","Create a managed volume for the artifacts and an external volume with LOCATION for the reference dataset.","Create an external volume for the artifacts so that dropping it purges their files, and a managed volume for the reference dataset."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00288","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:85eeb87755b11b80e6ce9f851071e7e5b1a35c81e232cf10d05fa431cf9b3b4e","aliases":[],"snapshot":{"statement":"The governance team granted the group marketing_analysts all of its access to catalog main with one set of statements:\n\n```sql\nGRANT USE CATALOG ON CATALOG main TO marketing_analysts;\nGRANT USE SCHEMA ON CATALOG main TO marketing_analysts;\nGRANT SELECT ON CATALOG main TO marketing_analysts;\n```\n\nThe group's project later ends. An administrator runs REVOKE SELECT ON CATALOG main FROM marketing_analysts, and the group holds no grants other than the three shown. What happens to its members' access to tables in main.sales and main.marketing?","options":["They lose SELECT on every table under main at once, because those reads were inherited from the catalog-level grant.","They keep reading the tables, because the surviving USE CATALOG and USE SCHEMA privileges are what authorize data access.","Their access persists until an administrator repeats the REVOKE on every schema and every table under main.","The REVOKE is ignored below the catalog, because Unity Catalog copies inherited privileges into each table when granting them."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00289","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e1f08c035b19436430fd90885a00446b460ec4ca7dda2429541d7f7bcbbf1328","aliases":[],"snapshot":{"statement":"During a security architecture review, the team is documenting every kind of object that Unity Catalog can protect with grants, beyond the familiar catalogs, schemas, and tables. The reviewer asks which other object types are securables that carry their own privilege types. Which answer is accurate?","options":["Only tables are securables; every other object is protected indirectly through the privileges of the catalog that contains it.","Views, volumes, and functions exist as governed objects, but none of them carries privileges of its own.","Storage credentials and external locations are cloud-side constructs that Unity Catalog references but cannot govern.","Views, volumes, functions, external locations, storage credentials, and connections are all securables with their own privileges."],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00290","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:17a64cea70f6404c04c13074fb70bb34d066ed6518ea6f9b2e705cf5856e3ae6","aliases":[],"snapshot":{"statement":"Five workspaces share one Unity Catalog metastore, and by default every workspace can see every catalog. A new compliance requirement says one analytics catalog may be exposed to only two of the five workspaces, without splitting it into separate copies.\n\nWhich mechanism implements this?","options":["Create a second metastore and replicate the whole catalog into it for the other workspaces","Grant USE CATALOG to the analysts group and rely on group membership per workspace","Convert the catalog's tables to external tables stored in a separate storage account","Workspace-catalog binding: assign the catalog to a restricted list of workspaces"],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00291","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:23b877be19db0dede9fc24db1d3104f8df1e8ed57c4ae826fc9f3bd1fb203bec","aliases":[],"snapshot":{"statement":"Security policy requires rotating the AWS IAM role that backs the company's lakehouse storage every quarter. The role is attached to a Unity Catalog storage credential that several external locations and external tables depend on. The platform team must implement the rotation without recreating objects or interrupting nightly pipelines. Which approach meets the requirement?","options":["Create a second storage credential, then drop and recreate every external location and table against it during a release window.","Update the existing storage credential with the new IAM role, so external locations and tables keep referencing the same object.","Pause all pipelines during a maintenance window, rotate the role in AWS, and wait for Unity Catalog to rediscover the credential.","Create a new storage credential every quarter, because credentials are immutable once an external location references them."],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00292","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7448266f7946efd348f614d74b7b57f5af5778462278381c4a55c2fe40f10903","aliases":[],"snapshot":{"statement":"A platform team keeps its production notebooks in a Databricks Git folder cloned from the team's GitHub repository. External auditors must be allowed to open and read the folder's notebooks and run them on sample data, but they must not edit files, create branches, or push changes. How should a workspace admin grant the auditors' group this access?","options":["Grant the auditors CAN RUN on the Git folder: workspace objects like Git folders use their own permission levels (CAN READ, CAN RUN, CAN EDIT, CAN MANAGE).","Grant the auditors SELECT on the Git folder through Unity Catalog, because Git folders are securables governed by the metastore like tables.","Grant the auditors nothing on the folder, because every workspace user can already view, edit, and run any Git folder by default.","Grant the auditors CAN RUN on the catalog's Delta tables, because permissions on a Git folder automatically extend to the data its notebooks read."],"answer":"A","reference":"https://docs.databricks.com/aws/en/repos/git-folders-concepts","referenceTitle":"Git folders concepts — Databricks Docs"}}
{"id":"RAN-DBDEA-00293","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:8144370819f2a75c89e49ca740ee5ee3e8219b9d8882e973c9181b6e7c2eaaac","aliases":[],"snapshot":{"statement":"A production pipeline reads its database password from the Databricks secret scope etl-secrets. The analysts' group must be able to read an API key stored in that same scope, but the group must not be allowed to change the scope's access rules or overwrite secret values. How should the admin grant the analysts this access?","options":["Grant nothing on the scope, because any workspace user can already read the secrets stored in every secret scope by default.","Grant the group SELECT on the secret scope through Unity Catalog, because secret scope ACLs are inherited automatically from Unity Catalog group privileges.","Grant the group the READ permission on the scope's ACL: READ allows reading secret values, and only MANAGE can change the scope's permissions.","Grant the group MANAGE on the scope, because MANAGE is the permission level that lets principals read secret values."],"answer":"C","reference":"https://docs.databricks.com/aws/en/security/secrets","referenceTitle":"Manage secrets — Databricks Docs"}}
{"id":"RAN-DBDEA-00294","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b5f3899bb07679c1d568a4dd6c2398a13630de1a9566c4d72ba48510d935924e","aliases":[],"snapshot":{"statement":"A contractor's engagement ends today, and the offboarding checklist requires that all access to Unity Catalog data be cut off before the end of the day. An access review confirms that the contractor's privileges on dozens of tables all arrive through membership in three groups; the contractor holds no direct grants. Which action cuts off the contractor's data access completely and verifiably?","options":["Revoke SELECT from each of the three groups on every table they can currently read, working through the dozens of objects one by one before the end of the day.","Wait for the contractor's personal access tokens to expire, which automatically disables the account's data access.","Rename the contractor's user account so that sign-in fails, and leave the group memberships and all grants exactly as they are.","Remove the contractor from the three groups: every privilege arrived through membership, so this single change cuts off all table access and is auditable."],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00295","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:98a2c34752c0802bd37e4991ef657a3006187e9f3c589e3e5b25acd6d984c5c2","aliases":[],"snapshot":{"statement":"A global company syncs its Databricks groups from its identity provider using automatic identity management. The group data-latam is a nested member of the corporate group data-global, which holds SELECT on the catalog analytics. A data engineer whose only group is data-latam runs a query against a table in analytics. What happens?","options":["The query fails: Databricks groups cannot be nested, so data-latam would have to be flattened into data-global before any grant can apply.","The query succeeds: members of nested groups inherit the parent's permissions, so the SELECT granted to data-global applies to data-latam's members.","The query fails: privileges granted to a group apply only to the users and service principals listed as its direct members.","The query succeeds, but only because Unity Catalog automatically duplicates the parent's grant onto every child group at the moment the nesting is created."],"answer":"B","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00296","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9f09fc8e25acbb07a8c4fa819ad676b0cc4be2e9dc1f1d85800ff1ca83d7a14d","aliases":[],"snapshot":{"statement":"Finance needs a monthly cost breakdown of Databricks usage by team for chargeback, computed from the lakehouse itself rather than from invoices or the cloud provider's console.\n\nWhich approach meets this?","options":["Query system.billing.usage, filtering by workspace and tags, aggregating DBUs and dollars per team","Parse the monthly PDF invoices with a notebook and load the results into a Delta table","Read the cloud provider's billing export, which already itemizes Databricks DBUs per team and workspace","Enable a workspace setting that emails each team leader their DBU total every month"],"answer":"A","reference":"https://docs.databricks.com/aws/en/admin/system-tables","referenceTitle":"System tables reference — Databricks Docs"}}
{"id":"RAN-DBDEA-00297","exam":"DBDEA","topics":["governance"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:4ea03f8da714b79fd9aac35bc35803e004604b9f0c29e47d5d4c8cb549ce9bbd","aliases":[],"snapshot":{"statement":"A company runs one Unity Catalog metastore shared by several business domains. Security policy requires that every managed table created in the finance catalog keeps its data files under a storage prefix dedicated to finance (s3://acme-managed/finance/), instead of the metastore root storage used by other domains. How should the platform team configure this?","options":["Spin up a second metastore whose root storage is the finance prefix, because managed storage locations exist only at the metastore level.","Leave the catalog as it is, because managed tables in every catalog are always written to the metastore's root storage location and can never use anything else.","Create the finance catalog with MANAGED LOCATION 's3://acme-managed/finance/', overriding the metastore root storage for that catalog's managed tables.","Create the finance tables as external tables with LOCATION under the finance prefix, because managed locations per catalog work only with external tables."],"answer":"C","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00298","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:1ee135bb50adee8dd6ad503d939e89b286098c5f7b06e0c127cd149aacaaa67f","aliases":[],"snapshot":{"statement":"During an architecture review, a new data engineer must explain how three Unity Catalog objects work together for cloud storage: the object that holds the IAM credential, the object that registers a storage path for external tables, and the object that gives notebooks governed access to files. Which explanation is correct?","options":["Volumes have replaced external locations entirely, so a single volume type now covers both the registration of storage paths for tables and practical access to files.","Storage credentials point to the catalogs that use them and carry the privilege to run queries against those catalogs.","External locations store the IAM credentials of every external table registered under them, one credential per table.","Storage credentials hold the IAM credential, external locations register a storage path referencing it, and volumes give governed access to files."],"answer":"D","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/","referenceTitle":"What is Unity Catalog? — Databricks Docs"}}
{"id":"RAN-DBDEA-00299","exam":"DBDEA","topics":["governance"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d8d0df5608493e1ad5e980fddde5acaa6f655033636cf67999488b3f27a70285","aliases":[],"snapshot":{"statement":"A Lakeflow Job writes manifest files directly to the cloud storage path s3://acme-dropzone/manifests/, which is registered in Unity Catalog as the external location acme_dropzone. The job runs under the service principal loader-sp and does not write through any table or volume. Which grant authorizes the job's writes?","options":["Grant the service principal WRITE FILES on the external location acme_dropzone, which authorizes direct writes on the registered path.","Grant the service principal SELECT on the external location acme_dropzone, which is the standard read privilege for governed data objects.","Grant the service principal WRITE VOLUME on the external location acme_dropzone, because volume privileges cover its registered storage paths.","Grant the service principal nothing, because Unity Catalog cannot authorize direct writes to a storage path outside tables and volumes."],"answer":"A","reference":"https://docs.databricks.com/aws/en/data-governance/unity-catalog/manage-privileges/","referenceTitle":"Manage privileges in Unity Catalog — Databricks Docs"}}
{"id":"RAN-DBDEA-00300","exam":"DBDEA","topics":["governance"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:fe0bbba98c773e02b33dba09f28100fafe49c8eb4bd86653243c438a7a32c817","aliases":[],"snapshot":{"statement":"During a security review, an auditor asks what the Databricks audit trail actually captures, expecting it to log only SQL queries. The platform lead must answer based on the documented coverage of the audit log system table (system.access.audit). Which answer is accurate?","options":["Only SELECT statements are captured; every other kind of event, including permission changes and object creation, is out of the audit trail's scope.","Only sign-in events are captured; queries run against the lakehouse are recorded solely in query history, never in audit logs.","A broad range of events is captured: logins and logouts, queries, permission changes such as grants, and object changes, each with the acting principal and timestamp.","Permission changes are not captured; audit events record queries and logins but never grants or revokes on governed objects."],"answer":"C","reference":"https://docs.databricks.com/aws/en/admin/system-tables/audit-logs","referenceTitle":"Audit log system table reference — Databricks Docs"}}
{"id":"RAN-AZ900-00001","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:c210fd032e1eb3f08840eca62d63067f2d9264ba0eb339cc90a866e721dbaa07","aliases":[],"snapshot":{"statement":"An IT director at a logistics company is preparing the team's first Azure project. During the kickoff meeting, a project manager asks for a plain-language definition of cloud computing.\n\nWhich statement should the IT director give?","options":["A model in which you buy physical servers up front and house them in the company's own datacenter","The delivery of computing services over the internet, including virtual machines, storage, databases, and networking","A way of delivering computing services over the company's private internal network, with no external provider involved","An arrangement where a vendor builds and operates a datacenter on the company's premises and bills a fixed annual fee"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/3-what-cloud-compute/","referenceTitle":"What is cloud computing — Microsoft Learn"}}
{"id":"RAN-AZ900-00002","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:211e77c89802fea0bf7d3e68f8489400e25adba1233071e252941460de31f416","aliases":[],"snapshot":{"statement":"A training nonprofit is onboarding a new cloud analyst. As a first exercise, the analyst lists what counts as a cloud computing service.\n\nWhich list is accurate?","options":["Virtual machines, storage, databases, and networking, plus newer offerings such as Internet of Things, machine learning, and AI","Only the physical hardware layer: the compute racks, power, and cooling that providers operate in their regions","Only ready-to-use business applications, such as email and customer relationship management tools","Only hosting plans for public websites, of the kind providers resell to small businesses"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/3-what-cloud-compute/","referenceTitle":"What is cloud computing — Microsoft Learn"}}
{"id":"RAN-AZ900-00003","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:45a67640885d8ab1457c7dd9d71ac429378e74844d409a4b7d7506bb0c41ea04","aliases":[],"snapshot":{"statement":"A streaming service is producing an interactive live event that will multiply its normal traffic for a few days; after the event, demand returns to its usual level. The team wants extra capacity for those days without waiting to build new datacenter capacity.\n\nWhich statement describes how cloud computing supports this?","options":["The team must buy enough hardware up front to cover the event's peak and keep it available for the rest of the year","The team uses the cloud to rapidly expand its IT footprint for the event and scale back down afterward, without building a new datacenter","The team's IT footprint stays constrained by the physical infrastructure of its own datacenter","The team rents a dedicated datacenter for a full year even though the extra capacity is needed for only a few days"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/3-what-cloud-compute/","referenceTitle":"What is cloud computing — Microsoft Learn"}}
{"id":"RAN-AZ900-00004","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b1f88947a8ccf952f6eabfa76a6f16fadea71383be5cae28f308e34bd552545b","aliases":[],"snapshot":{"statement":"A flower delivery company sees orders multiply around Valentine's Day and Mother's Day. In past years, preparing for those peaks meant long procurement cycles for servers, and much of that equipment sat idle between peaks.\n\nAt a fundamentals level, how does cloud computing change this planning?","options":["Planning still starts with long procurement cycles, but the provider finances the hardware purchases for the company","Planning is handled by the provider, which forecasts the company's demand and provisions capacity on its behalf","Planning becomes a single up-front decision: buy capacity for the highest expected peak and keep it running year-round","Planning shifts from long procurement cycles to on-demand provisioning: capacity is added for the peak and released afterward"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/3-what-cloud-compute/","referenceTitle":"What is cloud computing — Microsoft Learn"}}
{"id":"RAN-AZ900-00005","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ec0640c9bf384788ae766db3a383709db1ae02e20248b54f6304819d0872b2bc","aliases":[],"snapshot":{"statement":"A hospital network is moving its workloads to the cloud. A compliance officer asks which responsibilities always remain with the cloud provider, regardless of the service type the hospital uses.\n\nWhich statement is accurate?","options":["The information and data stored in the cloud, and the accounts that use the services","The physical datacenter, the physical network, the physical hosts, and the underlying infrastructure","The operating systems and the network controls, whichever service type is used","The applications and the devices allowed to connect to the cloud"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/4-describe-shared-responsibility-model/","referenceTitle":"Describe the shared responsibility model — Microsoft Learn"}}
{"id":"RAN-AZ900-00006","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:802d9783a504042c4653e3eb2fccc7df567949084fc35dd0830d48a220f374a3","aliases":[],"snapshot":{"statement":"A retailer has adopted a public cloud. A newly hired security engineer is reviewing what the company must keep doing itself, no matter which cloud service type it adopts.\n\nWhich responsibilities always stay with the retailer?","options":["The physical hosts, the physical network, and the underlying infrastructure that run its workloads in the provider's datacenter","The operating systems, the network controls, and the applications, whichever service type the retailer adopts","The power, the cooling, and the physical security of the datacenter where its workloads run","The data in the cloud, the devices allowed to connect, and the accounts and identities in the environment"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/4-describe-shared-responsibility-model/","referenceTitle":"Describe the shared responsibility model — Microsoft Learn"}}
{"id":"RAN-AZ900-00007","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:2df1593ccf5817ff597af8c4762dfa1ed12f09e6b446e76546a4d04bbcc47b48","aliases":[],"snapshot":{"statement":"An external auditor is comparing the company's former on-premises datacenter with its current public cloud setup, and asks how responsibilities shifted when the workloads moved.\n\nWhich statement is accurate?","options":["On-premises, the company handled everything; in the cloud, physical security, power, cooling, and connectivity moved to the provider","On-premises, the provider handled physical security; in the cloud, the company became responsible for everything","Moving to the cloud transferred responsibility for the stored data and for access security over to the provider","Moving to the cloud made the provider responsible for patching the company's applications in every service type"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/4-describe-shared-responsibility-model/","referenceTitle":"Describe the shared responsibility model — Microsoft Learn"}}
{"id":"RAN-AZ900-00008","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:80a5b8fa4c2ca8243d499e4a807984f848470d4ed76c4ea5c19b8c22f2f459a6","aliases":[],"snapshot":{"statement":"A manufacturing company's cloud architect is briefing the infrastructure team on how the shared responsibility model behaves across service types, ahead of the team's first migration wave.\n\nWhich statement is accurate?","options":["SaaS places the most responsibility on the consumer, while IaaS shifts nearly everything to the cloud provider","All three service types split responsibilities identically between the provider and the consumer, whatever the workload demands","IaaS places the most responsibility on the consumer, SaaS on the provider, and PaaS splits responsibility between the two","The service type has no bearing on who maintains the operating systems and the applications running in the cloud"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/4-describe-shared-responsibility-model/","referenceTitle":"Describe the shared responsibility model — Microsoft Learn"}}
{"id":"RAN-AZ900-00009","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:dc5c95e65669cc34985d04e32838fe4798bd7bb82cb2d5973317c7c0259b5ae3","aliases":[],"snapshot":{"statement":"A fintech runs two database workloads in the cloud. One is a SQL database that the team installed on its own virtual machines and patches itself. The other is a managed cloud SQL database service. Under the shared responsibility model, who maintains each database engine?","options":["The provider maintains both engines, because all database software running in the cloud is managed by the provider","The fintech maintains the engine on its virtual machines; the provider maintains the managed cloud SQL database","The fintech maintains both engines, because anything database-related always stays with the consumer","The provider maintains both engines, though the fintech stays responsible for the data in each case"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/4-describe-shared-responsibility-model/","referenceTitle":"Describe the shared responsibility model — Microsoft Learn"}}
{"id":"RAN-AZ900-00010","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:7408715dc68f18b04e696f923493e720063e38a14bb321aa8b348979a355f306","aliases":[],"snapshot":{"statement":"A university is evaluating its first cloud deployment, and a lecturer on the evaluation committee asks what a public cloud is.\n\nWhich statement is accurate?","options":["A cloud used by a single entity, with resources dedicated to one organization and no other tenants sharing them","A cloud that combines the university's on-premises datacenter with a provider's cloud in an interconnected environment","A cloud operated jointly by two providers, in which workloads can move freely between the providers' platforms","A cloud built, controlled, and maintained by a third-party provider, where anyone can purchase cloud services and use them"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00011","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:7028345383ef3dff70c0c67cc7d379599892064ce5682e1fcc3f8546b76a6b4d","aliases":[],"snapshot":{"statement":"A pharmaceutical company wants cloud technology, but its resources must be dedicated to the organization alone. A team member asks what a private cloud is.\n\nWhich statement is accurate?","options":["A cloud environment used by a single entity, hosted on-site or in a dedicated offsite datacenter, even a third-party one","A cloud where anyone can purchase access and where resources are shared with other tenants in the provider's own datacenter","A temporary public deployment used to handle surges in demand on top of the organization's own datacenter","A cloud model in which two public providers jointly run the environment as if it were a single cloud"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00012","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e3517daac20a63a21493075e5cb3ad05a04400de6ffacddd37169dcaad114f60","aliases":[],"snapshot":{"statement":"An insurance company keeps a private cloud in its own datacenter and connects it to a public cloud so it can absorb temporary demand spikes without changing its steady-state setup.\n\nWhich statement about this environment is accurate?","options":["It is two public clouds, from different providers, operated as one interconnected environment, with the private datacenter retired","It is a public cloud in which the provider dedicates hardware to a single tenant inside its own datacenter","It is a private cloud whose spare resources are shared with other tenants during peak periods to absorb the demand spikes","It is a hybrid cloud: interconnected public and private clouds, where the private side surges for temporary demand using public resources"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00013","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3fce528727e9b5ee2109d2c18a247921b6c200e2e7328d9644693c34341cbf23","aliases":[],"snapshot":{"statement":"A gaming studio uses one public provider for its development platform and is moving part of its workloads to a second public provider to use features the first one lacks, while the migration is still in progress.\n\nWhich statement describes this environment accurately?","options":["It is a hybrid cloud scenario, because the studio is combining two different cloud models into one","It is a multicloud scenario: the studio uses two public cloud providers and manages resources and security in both environments","It is a private cloud scenario, because each provider dedicates its environment to the studio alone","It is still a single public cloud scenario, because using several public providers at once does not change the model"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00014","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:06821b828de9587cb0d57e4a9a6d852c5138aa89d2b38eb1fb6083f61fa98e6f","aliases":[],"snapshot":{"statement":"A solutions architect is preparing a briefing that compares cloud models for a client's board. One slide pairs characteristics with the models, and the architect checks it before the meeting.\n\nWhich pairing is accurate?","options":["Private cloud: no capital expenditures to scale up, because the provider pays for the hardware","Hybrid cloud: you have complete control over resources and security at all times","Public cloud: no capital expenditures to scale up — you pay only for what you use","Public cloud: your data isn't collocated with other tenants' data"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00015","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:3ec23416da14c2bec0be9e5ad65ee572be62a8d700f83fe0245b19997a5be263","aliases":[],"snapshot":{"statement":"During a proposal review, a client asks the architect to confirm which statement about cloud models is accurate before it goes into the final deck.\n\nWhich statement is it?","options":["In a private cloud, hardware must be purchased for startup and maintenance, and the organization maintains and updates it","In a public cloud, the consumer purchases and maintains the hardware that runs the services in the provider's datacenter","In a hybrid cloud, the provider decides where each application runs, based on cost optimization","In a multicloud environment, the security policies of one provider automatically govern the other provider's resources"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00016","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f411691307f39882fa90d0b48c182580968fce534a9876ed22a27ea6774608d9","aliases":[],"snapshot":{"statement":"A defense contractor must run all of its workloads on dedicated infrastructure that the organization controls. Its data cannot be collocated with other tenants' data, and it will not use any public cloud resources for these workloads.\n\nWhich cloud model fits these requirements?","options":["Public cloud, because dedicated hardware can be reserved in public regions for use by a single tenant","Hybrid cloud, because the sensitive workloads stay on the private side of the environment","Private cloud, because resources are dedicated to one entity and data isn't collocated with other tenants' data","Multicloud, because spreading the workloads across two public providers isolates them from other tenants"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00017","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:3b5b9c1344e668a8ac83db437d27c4e46e017ac2664b3a580a9bf5becf524342","aliases":[],"snapshot":{"statement":"Two founders are building a mobile app with no capital to buy servers. Usage is expected to grow unpredictably if the app takes off. They want to pay only for what they use, provision and deprovision applications quickly, and avoid managing resources in multiple environments.\n\nWhich cloud model fits best?","options":["Public cloud: no capital expenditures to scale up, quick provisioning, and you pay only for what you use","Private cloud: hardware purchased up front for startup and maintenance, in exchange for complete control over resources and security","Hybrid cloud: start in a connected on-premises datacenter and add public capacity later when demand grows","Multicloud: split the app across two public providers from day one, to spread the load"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00018","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:8470e837ce68bdd358f2df3642685a52a50ae88c57d4e9e4f0c2a0a3198c23cc","aliases":[],"snapshot":{"statement":"A tax-preparation firm runs a private cloud that carries a steady load for most of the year. During filing season, demand rises sharply for a few weeks. The firm wants to keep its steady-state workloads where they are and temporarily deploy extra capacity from a public cloud, connected to its own environment.\n\nWhich approach does this describe?","options":["Public cloud: migrating the whole private environment to the provider for the season and keeping it there","Private cloud: buying additional hardware for the peak and keeping it available for the rest of the year","Multicloud: renting extra capacity from a second public provider during the season, alongside the first","Hybrid cloud: letting the private cloud surge for increased, temporary demand by deploying public cloud resources"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00019","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:30665119b3341713e40734b693f36d41e5b933967c3292ad8bd8c066192ddc50","aliases":[],"snapshot":{"statement":"An energy utility uses one public provider for its customer portal and a second public provider for analytics, because each offers features the other lacks; the migration between them is still in progress.\n\nBeyond naming the model, what does operating this environment require?","options":["Extending one provider's security policies automatically across the other provider's resources","Managing resources and security in both public cloud environments","Consolidating both providers into a single public cloud before management becomes possible","Managing only the environment that hosts the newest workloads"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00020","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:98fb617a222fa2135fe6dcd4301e3dcb56e5a4566904323b293a1d230ed63b9d","aliases":[],"snapshot":{"statement":"A national research lab must keep its most sensitive workloads on infrastructure it controls, but it also wants public cloud capacity for temporary surges and for collaboration tools. It needs to decide, per workload, where each application runs, and it must retain control over security, compliance, and legal requirements.\n\nWhich cloud model supports all of these requirements?","options":["Public cloud, with the provider deciding where each workload runs in order to optimize costs","Private cloud, which offers complete control but no way to add public capacity for temporary surges","Hybrid cloud, which provides the most flexibility: you determine where applications run and control security and compliance","Multicloud, which keeps the sensitive workloads isolated by running them across two public providers"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/5-define-cloud-models/","referenceTitle":"Define cloud models — Microsoft Learn"}}
{"id":"RAN-AZ900-00021","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:a6812eeca2a1838f265bb263052f58de1b15dbb2cb63c6af5daa16899db74f1d","aliases":[],"snapshot":{"statement":"The CFO of a mid-sized company asks the cloud team to explain the consumption-based model in one sentence, so that she can repeat it to the board without notes.\n\nWhich statement is accurate?","options":["You pay for the IT resources you use and nothing more, renting compute and storage and releasing them when done","You pay a fixed annual fee that covers unlimited use of all the resources you request","You buy the hardware up front and then pay only for the electricity it consumes in the company's own datacenter","You pay in advance for the capacity you expect to consume over the coming three years"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/6-describe-consumption-based-model/","referenceTitle":"Describe the consumption-based model — Microsoft Learn"}}
{"id":"RAN-AZ900-00022","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:9853c39fb4c2200d333ef8c867b6701124a72b8c0fe3ae6f0f65b6d587a79a8b","aliases":[],"snapshot":{"statement":"An accountant is classifying the company's IT spending after its move to the cloud, using the CapEx/OpEx distinction from cloud budgeting.\n\nHow is spending on cloud services classified, and why?","options":["As capital expenditure, because cloud capacity is purchased up front like datacenter hardware","As capital expenditure, because the provider buys the hardware and passes the cost on in monthly bills","As operational expenditure, because you pay for cloud services as you consume them over time","As operational expenditure, because you buy the servers and pay for them in monthly installments"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/6-describe-consumption-based-model/","referenceTitle":"Describe the consumption-based model — Microsoft Learn"}}
{"id":"RAN-AZ900-00023","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3c789523665ca872172ce57a0f1657327b0c6536bdba978fe72f0fffe18c5e70","aliases":[],"snapshot":{"statement":"The head of IT planning at a manufacturer is explaining to the board why traditional datacenter capacity planning is hard.\n\nWhich pair of outcomes describes the risks of estimating resource needs in advance in a traditional datacenter?","options":["Overestimate and applications suffer degraded performance; underestimate and you overspend on idle infrastructure","Overestimate and you overspend on idle infrastructure; underestimate and the provider ships additional hardware overnight","Overestimate and you overspend on infrastructure that sits idle; underestimate and your applications suffer degraded performance","Overestimate and you still pay only for what you use; underestimate and performance is unaffected"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/6-describe-consumption-based-model/","referenceTitle":"Describe the consumption-based model — Microsoft Learn"}}
{"id":"RAN-AZ900-00024","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:1af9fac4041301b2994b578032c86f2ff0eaf7b011c1f187856dd5491077f000","aliases":[],"snapshot":{"statement":"A nonprofit runs donation drives whose traffic is hard to predict, and its budget is tight. The cloud team lists the benefits of the consumption-based model for the board.\n\nWhich list is accurate?","options":["Up-front hardware purchases and fixed capacity, with spending that stays stable regardless of how demand moves","The provider absorbs the full cost of any cloud capacity that you reserve in advance but then leave sitting idle","The ability to add resources when demand increases, though releasing them later takes weeks of manual processing","No upfront costs, no capacity purchased ahead of need, and the ability to add and release resources as demand changes"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/6-describe-consumption-based-model/","referenceTitle":"Describe the consumption-based model — Microsoft Learn"}}
{"id":"RAN-AZ900-00025","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:5a9fe4705f9d9e10b7164e89e6b87f882624f14843c9228327fa20952a2991f3","aliases":[],"snapshot":{"statement":"A company is closing its aging datacenter and moving entirely to the cloud. Its finance lead must classify each cost as CapEx or OpEx, using the definitions from cloud budgeting.\n\nWhich classification is accurate?","options":["Buying servers, network hardware, and datacenter space is CapEx; the monthly bills for cloud services consumed are OpEx","Buying servers, network hardware, and datacenter space is OpEx; renting cloud capacity is CapEx","Both the hardware purchases and the cloud bills are CapEx, since both are IT costs","Both the hardware purchases and the cloud bills are OpEx, since both are paid over time"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/6-describe-consumption-based-model/","referenceTitle":"Describe the consumption-based model — Microsoft Learn"}}
{"id":"RAN-AZ900-00026","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:5a194a96c77e55762cc4279ab49233a63d8ee3070f5b8dd919222ee5a900219a","aliases":[],"snapshot":{"statement":"A small design agency wants cloud services with no upfront commitment. Its workload is unpredictable, so the team wants to pay only for what it uses, to be able to cancel anytime, and to keep the services running for exactly as long as the team itself chooses.\n\nWhich pricing approach matches these requirements?","options":["Reserved instances: a one- or three-year commitment in exchange for a discount","Spot capacity: unused capacity at a discount, which Azure reclaims whenever it needs the capacity back","Pay-as-you-go: you pay only for the services you consume, with no upfront commitment","A fixed monthly fee that covers unlimited use of the services"],"answer":"C","reference":"https://azure.microsoft.com/en-us/pricing/purchase-options/pay-as-you-go/","referenceTitle":"Pay as you go — Microsoft Azure"}}
{"id":"RAN-AZ900-00027","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ee4cf8f93b5930885d6db15a56eb3925f0655703253136cfa308ddc4c825e197","aliases":[],"snapshot":{"statement":"A logistics company runs a shipment-tracking system that uses roughly the same amount of compute around the clock, every day, for the foreseeable future. The cloud bill for that system is the team's largest cost line.\n\nWhich pricing approach is most likely to reduce these costs, and why?","options":["Pay-as-you-go, because its per-hour rates are the lowest for workloads that run all day and night","Azure Reservations, because a one-year or three-year commitment earns a discount on the matching resources","Spot capacity, because it offers the deepest discounts and steady workloads are never evicted","A reservation purchased in a different region, because reservation discounts apply across regions automatically"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/cost-management-billing/reservations/save-compute-costs-reservations","referenceTitle":"What are Azure Reservations? — Microsoft Learn"}}
{"id":"RAN-AZ900-00028","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6ac6480de03d4961272bad78153d062851a78fb9f27f7aa25d91b6d99cdee708","aliases":[],"snapshot":{"statement":"A film studio runs nightly rendering jobs that can be restarted if they are interrupted, and it keeps a few dev and test environments that tolerate downtime. It wants the lowest possible compute prices.\n\nWhich pricing approach fits, and what is the trade-off?","options":["Azure Reservations, because rendering jobs qualify for the deepest discounts without any commitment","Pay-as-you-go, because it guarantees capacity is available whenever demand rises and prices stay flat","A free monthly allowance, because compute used for rendering is always free up to a monthly cap","Spot capacity: unused capacity at significant savings, accepting that Azure can evict the workloads when it needs them back"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/virtual-machines/spot-vms","referenceTitle":"Azure Spot Virtual Machines — Microsoft Learn"}}
{"id":"RAN-AZ900-00029","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:8b77a4066855512e5490c202772452f480839a98097c61ba33b4edc517766fed","aliases":[],"snapshot":{"statement":"A cloud architect is briefing a client that has heard of three ways to pay for compute: committing for a discount, buying unused capacity, and paying for what is used.\n\nWhich statement correctly distinguishes the three pricing models?","options":["Reservations trade a long-term commitment for lower prices; spot trades tolerance for interruptions for significant savings; pay-as-you-go charges only for what you consume","Reservations trade tolerance for interruptions for lower prices; spot trades a long commitment for the deepest savings; pay-as-you-go bills a fixed annual fee","Reservations offer the lowest prices with no commitment; spot offers reserved capacity at fixed prices; pay-as-you-go bills monthly for unlimited use","Reservations apply only to storage charges; spot applies only to database workloads; pay-as-you-go applies only to networking costs"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/cost-management-billing/reservations/save-compute-costs-reservations","referenceTitle":"What are Azure Reservations? — Microsoft Learn"}}
{"id":"RAN-AZ900-00030","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6fd124bc8be8d10090aa8451257c0abe58a88f8c3bf319c6919240314833416a","aliases":[],"snapshot":{"statement":"An online magazine's traffic spikes unpredictably when stories go viral. The publishing system cannot tolerate interruptions, and the team wants to avoid long-term commitments while demand is still uncertain.\n\nWhich pricing approach fits best?","options":["Spot capacity, because evictions only occur when traffic is low anyway","Pay-as-you-go: you pay only for what you consume, and costs scale as workload needs change","Azure Reservations, because a three-year commitment is best while demand is still uncertain","A single annual bill that averages out the viral peaks across the whole year"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-compute/6-describe-consumption-based-model/","referenceTitle":"Describe the consumption-based model — Microsoft Learn"}}
{"id":"RAN-AZ900-00031","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:15e836bbd24d7ca09ee4023fbf4842bb82066142cac6029f9c898be68419ac81","aliases":[],"snapshot":{"statement":"A junior developer on the team asks what \"serverless\" actually means in Azure, after hearing the term in a design review.\n\nWhich statement is accurate?","options":["It means there are no servers involved anywhere in the process of building and running your code","It means your code runs only on your own on-premises hardware, inside the company's own datacenter","It's a solution like Azure Functions: you write the code, and the cloud provides the resources to keep it running","It means servers are rented by the hour and must be kept running between requests, even when idle or unused"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-functions/functions-overview","referenceTitle":"Azure Functions overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00032","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:85daefc75067599985f4ddf21650b7742b53cb962308f3b47a5040a823227987","aliases":[],"snapshot":{"statement":"A photo-sharing app wants to generate thumbnails automatically whenever a user uploads a picture to blob storage — with no polling loop and no always-on server.\n\nWhich Azure Functions capability supports this?","options":["Event-driven triggers and bindings that connect the function to other services without extra code","Scheduled timers that fire at fixed nightly intervals, whether or not uploads happened","A manually started console that an operator opens whenever new uploads arrive","A dedicated virtual machine that listens continuously for new uploads to process"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-functions/functions-overview","referenceTitle":"Azure Functions overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00033","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:bfa40a0c51d24504f18a2133ed587e6b51a1a1a67750ae3df8d747e1df772d35","aliases":[],"snapshot":{"statement":"A small product team wants to add a nightly data-cleanup job to its service, but it has no one available to deploy, patch, or maintain servers for it.\n\nWhich statement describes how serverless computing helps in this situation?","options":["The team ships the code together with a server image that it keeps patched and updated itself","The team writes the code, and an external contractor deploys and maintains all the underlying servers for it","The code runs whenever a team member logs in and starts the nightly job by hand","The team provides the code, and the cloud provides the resources — no servers to deploy or maintain"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/azure-functions/functions-overview","referenceTitle":"Azure Functions overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00034","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:9c115db4b990978d43e832a20e024fc002325d22e0fe982abb283c4ca6b09713","aliases":[],"snapshot":{"statement":"A team is choosing between a classic always-on hosting plan and a serverless approach for a web API that mostly sits idle and wakes up for short bursts.\n\nWhich statement correctly distinguishes the serverless approach in this comparison?","options":["It requires at least one always-warm instance running at all times so that responses are never delayed by cold starts","Functions are invoked by HTTP triggers, scale fast on events, and bill pay-as-you-go — not dedicated instances running around the clock","Its functions must run continuously in the background, polling for incoming requests to react","It always costs more than a dedicated plan, because idle time is billed at the full hourly rate"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-functions/functions-overview","referenceTitle":"Azure Functions overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00035","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a118100aeb8e863a5eade265c9dca1b4b64361a52593a6203dab70dbdf573137","aliases":[],"snapshot":{"statement":"An IoT startup needs to capture and transform data from device event streams on the way to storage, reacting to each event as it arrives, without managing any servers.\n\nWhich approach fits, and why?","options":["A fleet of virtual machines that polls the device queue on a fixed interval, day and night","A nightly batch job that runs once a day, regardless of when the events actually arrive","Azure Functions with event-driven triggers: code runs in response to events, and the cloud keeps it running","A manually operated export tool that an engineer runs whenever the data is needed"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-functions/functions-overview","referenceTitle":"Azure Functions overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00036","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:40bf0a44a035dfe01f22d42adee90b9c878a248f112c7e8330e31191698453c9","aliases":[],"snapshot":{"statement":"A regional hospital network is moving its patient scheduling application to Azure. The CIO's top requirement is that the application stays accessible to clinicians whenever they need it, even when individual components or the underlying infrastructure experience disruptions. Which cloud benefit is the CIO describing?","options":["Scalability, because more resources can be added whenever the volume of user traffic increases","High availability, which focuses on ensuring maximum availability regardless of disruptions or events","Cost predictability, because resource usage can be tracked in real time","Elasticity, because capacity automatically expands and contracts to match live demand"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/2-high-availability-scalability-cloud","referenceTitle":"Describe the benefits of high availability and scalability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00037","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:9a44b51172d8d1c6f4c59890fc7684c5acabedd007ee0e45d89bb0eea526c495","aliases":[],"snapshot":{"statement":"An engineering consultancy runs its 3D rendering workload on a single Azure virtual machine. Render jobs are taking too long, so the team decides to add more CPUs and more RAM to that same virtual machine. In cloud terminology, what is the team doing?","options":["Scaling out horizontally by adding more virtual machines to share the rendering jobs","Scaling in by removing rendering capacity until the next job arrives","Relying on autoscaling, since the platform adjusts the machine size on its own","Scaling up vertically by increasing the capabilities of an existing resource"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/2-high-availability-scalability-cloud","referenceTitle":"Describe the benefits of high availability and scalability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00038","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3d81b42f47cda845784921b19bf4ae4cb1c66e80953a9bb97ba5b073b731a0d0","aliases":[],"snapshot":{"statement":"A manufacturer's production planning application was not designed to run as multiple parallel instances; its architecture assumes one instance with shared in-process state. Month-end processing pushes that single instance to its limits for a few hours each month. Which approach should the cloud team take during that window?","options":["Vertically scale up the virtual machine during month-end processing, then scale it back down afterward","Horizontally scale out by adding extra application instances behind a load balancer","Redeploy the application across several regions so each region handles part of the workload","Configure automatic scaling rules so the platform adds instances when CPU usage spikes"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/2-high-availability-scalability-cloud","referenceTitle":"Describe the benefits of high availability and scalability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00039","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:9c4fd88252f1b23af877976f373834df40c40a58b11697fedaf74a94824e2bad","aliases":[],"snapshot":{"statement":"An online retailer is preparing a 48-hour flash sale. To absorb the surge, the web operations team plans to add several virtual machine instances to the website tier for the weekend and remove the extra instances once the sale ends. Which scaling pattern does this plan describe?","options":["Vertical scaling up, because each virtual machine receives more CPUs and RAM","Reliability, because the website quickly recovers from any component failures during the sale","Horizontal scaling out and then in, because the number of instances follows demand","Cost predictability, because the team forecasts the sale's cloud spending in advance"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/2-high-availability-scalability-cloud","referenceTitle":"Describe the benefits of high availability and scalability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00040","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0be15b69e725073d710a805f552648ea1c2585e2f0c0113be9a03a75826d3e65","aliases":[],"snapshot":{"statement":"An education technology platform serves steady traffic most of the year, with extreme spikes during course registration weeks. On-premises, the company had to buy hardware sized for the peak that then sat idle. After moving to Azure, which pairing of cloud benefits most directly resolves both the capacity spikes and the wasted idle capacity?","options":["Scalability with consumption-based billing: grow for peaks, shrink afterward, and pay only for what you use","High availability with service-level agreements: the uptime guarantees keep the platform online during the registration peaks","Reliability with a decentralized design: resources deployed in additional regions absorb the registration traffic","Governance with cloud-based auditing: regular compliance checks verify that resources match demand patterns"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/2-high-availability-scalability-cloud","referenceTitle":"Describe the benefits of high availability and scalability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00041","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:18ac0f52bb768a0cfe4ee59a2036dca470f7ec08f2fe87c1c3f345e4f0923531","aliases":[],"snapshot":{"statement":"A ticketing platform sees demand explode within minutes when popular events go on sale, and collapse just as quickly afterward. Operations wants capacity to follow this demand automatically — growing and shrinking without anyone adding or removing servers manually — so that performance stays consistent and the company does not pay for idle capacity. Which concept does the operations team want?","options":["Vertical scaling, because upgrading the size of each server absorbs the higher load","Horizontal scaling, because operations staff can add and remove virtual machines as traffic changes","High availability, because service-level agreements keep the service online during spikes","Elasticity, because resources automatically expand and contract to match real-time demand"],"answer":"D","reference":"https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-is-elastic-computing","referenceTitle":"What is elasticity in cloud computing? — Microsoft Azure"}}
{"id":"RAN-AZ900-00042","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7eaa8716afa64e65662d83a7cd409bb9efcfe61fca34ec49df3c5517c474b203","aliases":[],"snapshot":{"statement":"A logistics company is selecting Azure services for its shipment tracking platform. Before signing off, the legal team wants to review Microsoft's formal commitments about how much uptime each Azure service guarantees. Where should the cloud team point the legal team?","options":["The total cost of ownership (TCO) calculator, which compares on-premises and cloud operating costs","The service-level agreements (SLAs), which state the uptime guarantees that depend on the service","The Azure Well-Architected Framework, which provides design guidance for reliable workloads","The Azure Pricing Calculator, which estimates the monthly cost of each service"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/2-high-availability-scalability-cloud","referenceTitle":"Describe the benefits of high availability and scalability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00043","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:f2ac4b349df64078652fdf38497b341603d1b56829c28962e1851b272f496bb7","aliases":[],"snapshot":{"statement":"An insurance company is designing its claims processing platform on Azure. The lead architect summarizes one design goal this way: \"If any component fails, the system must be able to recover from failures and continue to function.\" Which cloud benefit is the architect describing?","options":["High availability, which focuses on maximizing uptime regardless of disruptions or outages","Scalability, which adjusts the amount of resources to meet demand","Reliability, a system's ability to recover from failures and keep functioning","Cost predictability, which forecasts future cloud spending from usage patterns"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/3-reliability-predictability-cloud","referenceTitle":"Describe the benefits of reliability and predictability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00044","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b30688295b3f7eda1d9a395121e8775b7d1a05f638cc782ea0381b1e7f0ca419","aliases":[],"snapshot":{"statement":"A multinational utility company runs a grid monitoring API that must keep serving requests even if a catastrophic event, such as a natural disaster, takes an entire Azure region offline. When the architect explains that users would be served from other regions, which characteristic of the cloud is the architect counting on?","options":["The decentralized design, which spreads resources across regions so others keep running when one fails","The centralized design, which concentrates all resources in a single, highly secured datacenter","The consumption-based model, which charges you only for the capacity actually used in each region","Automatic patching, which keeps every regional deployment updated and consistent over time"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/3-reliability-predictability-cloud","referenceTitle":"Describe the benefits of reliability and predictability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00045","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:decf1290a56982141536afcfe49c7095c1eb7441705bc14c20a415949f493700","aliases":[],"snapshot":{"statement":"A video streaming startup is planning its growth and wants playback quality and response times to stay consistent as viewer numbers fluctuate through the day. The product owner asks which cloud benefit focuses on predicting the resources needed to deliver that positive experience. What is the correct answer?","options":["Cost predictability, which forecasts cloud spending from usage data and trend analytics","Performance predictability, which focuses on resources needed for a positive customer experience","Reliability, which recovers the platform from failures and keeps it functioning","Governance, which enforces corporate standards and requirements on deployed resources"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/3-reliability-predictability-cloud","referenceTitle":"Describe the benefits of reliability and predictability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00046","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b19d1c6ee40182cf1d56cbe980f6fce46e24b6f6dd706d774dd37c928daf1758","aliases":[],"snapshot":{"statement":"A fintech's finance team must submit a quarterly cloud budget. They want to track resource usage in real time, apply data analytics to find spending patterns, and get an estimate of potential spend for a new service before it is deployed. Which benefit and tool pairing should the team use?","options":["Performance predictability with load balancing, which redirects traffic to less stressed areas","Reliability with a decentralized design, which deploys workloads across multiple regions","High availability with service-level agreements, which document each service's uptime guarantee","Cost predictability with the Azure Pricing Calculator, which estimates potential cloud spend"],"answer":"D","reference":"https://azure.microsoft.com/en-us/pricing/calculator/","referenceTitle":"Azure Pricing Calculator — Microsoft Azure"}}
{"id":"RAN-AZ900-00047","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:dd1c9041b9f7a1d2eb08c59de611036991185ee1a71b84983ee89677ffc206a2","aliases":[],"snapshot":{"statement":"An online learning platform enrolls a wave of new students at the start of each semester. The operations team configures rules so that additional resources deploy automatically when demand rises and scale back automatically when it drops. How does autoscaling support the platform's performance predictability?","options":["It deploys additional resources when demand rises and scales back when the demand drops","It redirects traffic from overloaded instances to less stressed areas of the deployment","It guarantees a fixed monthly spend by keeping the resource count constant","It raises each service's uptime guarantee during periods of peak demand"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/3-reliability-predictability-cloud","referenceTitle":"Describe the benefits of reliability and predictability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00048","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:9f2f0dbce66295d420b85548bf401a4a698e2573632d1391b598edbcf212d2ca","aliases":[],"snapshot":{"statement":"During a cloud adoption workshop, the IT steering committee of a media company asks a solution architect to confirm several claims they have read about cloud reliability and predictability before approving the migration plan. Which claim is accurate?","options":["Reliability depends on a centralized design that concentrates workloads in a single primary datacenter.","Cost predictability is achieved primarily by scaling out more virtual machines whenever demand spikes.","Performance predictability focuses on predicting the resources needed to deliver a positive experience for customers.","Service-level agreements are the primary tool for forecasting future spending on cloud services."],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/3-reliability-predictability-cloud","referenceTitle":"Describe the benefits of reliability and predictability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00049","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:8cb6a12691d8f464a8118b18ddcd4bb618d5841fe87d36dc6d2afc18da65b649","aliases":[],"snapshot":{"statement":"A nonprofit with a two-person IT team is moving its donor management application to the cloud. The team wants the cloud provider to handle operating system patches and platform maintenance automatically so they can focus on the application itself. Which service model strategy meets this need?","options":["Infrastructure as a service, which keeps the team fully responsible for operating systems and patches","Platform as a service or software as a service, where patches and maintenance may be applied automatically","A hybrid cloud deployment, which preserves the team's current manual patching practices unchanged","Infrastructure as a service with automated backups, because regular backups keep the operating system current"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/4-security-governance-cloud","referenceTitle":"Describe the benefits of security and governance in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00050","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:5e48513083ebc4047f37a0ed233dd2ee09338168857fe8c296fa45aa0924bd29","aliases":[],"snapshot":{"statement":"A pharmaceutical company's security policy requires its internal team to keep full control of the operating systems, installed software, and patch timing on every server that runs a laboratory reporting application. Which cloud service type should the team choose to preserve this level of control?","options":["Software as a service, which transfers all operating system responsibilities to the provider","Infrastructure as a service, where the team manages the operating systems and installed software","Platform as a service, where patches are applied automatically on the provider's own schedule","Serverless computing, which runs code on demand without exposing any server-level configuration"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/4-security-governance-cloud","referenceTitle":"Describe the benefits of security and governance in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00051","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0cf40178321ec0782232fddcd5eed4917c8ffdcc9a282723ca6a6a8f76e6e2ac","aliases":[],"snapshot":{"statement":"A public sector agency learns that a new regulation requires a revised data retention configuration on every cloud resource it has deployed. The agency needs to roll the change out efficiently across all deployed resources, without administrators reconfiguring each one by hand. Which governance benefit of the cloud should the agency rely on?","options":["Vertical scaling, which increases the capabilities of each resource so changes apply faster","Cloud-based auditing, which identifies resources that no longer comply with the new standard","High availability, which keeps the agency's services online while the standard is applied","The ability to update resources at scale as standards and requirements change"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/4-security-governance-cloud","referenceTitle":"Describe the benefits of security and governance in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00052","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:947d00ec7d1d6dcb3a903aa372dc909b15ae9715b392c4c6765fea3570d415c3","aliases":[],"snapshot":{"statement":"An online news publisher is a frequent target of distributed denial-of-service (DDoS) attacks that can overwhelm its on-premises web servers, especially on election nights. The publisher is evaluating a move to Azure. Which security benefit of the cloud directly addresses this specific threat?","options":["Automatic operating system patching, which closes the software vulnerabilities exploited by the attackers","Deployment templates, which ensure that every server follows the same technical standards","Cloud providers are typically well suited to handle DDoS attacks, making networks robust","Consumption-based billing, which reduces the financial cost of absorbing the attack traffic"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/4-security-governance-cloud","referenceTitle":"Describe the benefits of security and governance in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00053","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:a51d69413288e4d6c81500c527a4a464c5d28058d0162a0316479f81f0f9abe2","aliases":[],"snapshot":{"statement":"A retail chain's compliance officer wants to know whether any of the company's deployed cloud resources have drifted away from its corporate security baseline, and wants suggestions for fixing the problems. Which governance capability of the cloud provides this?","options":["Automatic patching, which applies the latest software updates to deployed resources","Cloud-based auditing, which flags resources out of compliance and provides mitigation strategies","Automatic alerts based on configured metrics, which report resource performance in real time","Elasticity, which adjusts the number of deployed resources to match demand"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/4-security-governance-cloud","referenceTitle":"Describe the benefits of security and governance in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00054","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:212419bc022a587e71b7d2923a2c2bb17e351691ac01a0eb66b071d09acb318d","aliases":[],"snapshot":{"statement":"The IT steering committee of a manufacturing group is reviewing statements about security and governance in the cloud before approving a migration to Azure. Which statement should the solution architect confirm as accurate?","options":["Establishing a good governance footprint early keeps the group's cloud footprint updated, secure, and well managed.","Cloud governance tools replace the need to define corporate standards and regulatory requirements.","Governance features in the cloud apply only to platform as a service deployments.","Security responsibilities are identical across infrastructure as a service, platform as a service, and software as a service."],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/4-security-governance-cloud","referenceTitle":"Describe the benefits of security and governance in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00055","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ce2fe19097f10f4e9c8057916285bf03f16c6a12e62bbac328c08d121bf69c28","aliases":[],"snapshot":{"statement":"A newly hired cloud administrator at a university asks which interfaces are available for day-to-day management of the institution's cloud resources. Teammates prefer different tools: some like working in a browser, others automate everything from a shell. Which set of interfaces should the administrator mention?","options":["The cloud benefits of high availability, scalability, reliability, and elasticity","The TCO calculator, the pricing calculator, and the service-level agreements","Infrastructure as a service, platform as a service, and software as a service","A web portal, a command-line interface, APIs, and PowerShell scripts"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/5-manageability-cloud","referenceTitle":"Describe the benefits of manageability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00056","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d0afe5f05467da34519c65729e90b631883026d2c617537824d8bc4220936d52","aliases":[],"snapshot":{"statement":"A logistics company's small operations team wants the cloud to watch the health of its resources continuously and, when a resource starts failing, to replace it without anyone having to intervene. Which manageability capability of the cloud matches this requirement?","options":["Operators running PowerShell scripts to restart services by hand","Deploying new resources quickly from a preconfigured template","Monitoring the health of resources and automatically replacing failing ones","Receiving automatic alerts based on configured performance metrics"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/5-manageability-cloud","referenceTitle":"Describe the benefits of manageability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00057","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:4033145b4b9082c5e2db4509a551a84ad58126af509d7cdd494669b7ae8c2be6","aliases":[],"snapshot":{"statement":"A food delivery platform's operations lead wants to hear about performance problems as they happen — for example, an automatic notification whenever a configured metric crosses a defined threshold. Which manageability capability of the cloud provides this?","options":["Receiving automatic alerts based on configured metrics, for real-time awareness","Vertically scaling up the affected resource to add more capacity","A service-level agreement with a higher uptime guarantee for the service","A monthly cloud-based auditing report on resource compliance"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/5-manageability-cloud","referenceTitle":"Describe the benefits of manageability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00058","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ae94b81ba472d65689a690d9fbf95e7120f3d2bd4d6d065a5ffbc858d48bba2d","aliases":[],"snapshot":{"statement":"A nongovernmental organization runs its fundraising platform with a two-person IT team. They provision new environments from preconfigured templates, keep an eye on resource health in the web portal, and automate recurring cleanup tasks with a command-line script. Which cloud benefit does this working style illustrate?","options":["Manageability, because resources can be managed through the portal, CLI, and automation","Scalability, because resources are automatically added or removed as the demand changes","Reliability, because the workload recovers automatically from regional service failures","Cost predictability, because future spending is forecast from historical usage patterns"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/5-manageability-cloud","referenceTitle":"Describe the benefits of manageability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00059","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:1c148af93de844331b97b23cdf026bdc730f98173af8724a7c720467c3eb8e47","aliases":[],"snapshot":{"statement":"An airline's cloud engineering team is reviewing the two aspects of manageability described in Azure fundamentals training. The team lead asks each engineer to classify a capability as management of the cloud or management in the cloud. Which capability is correctly classified as management in the cloud?","options":["Automatically scaling resource deployment based on need","Automatically replacing failing resources detected by health monitoring","Managing the cloud environment and its resources by using PowerShell","Receiving automatic alerts on resources based on configured metrics"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/5-manageability-cloud","referenceTitle":"Describe the benefits of manageability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00060","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7a9569de82f66cfa97d7829d8477ff657ce2e5a18f2617a590958292644a06c3","aliases":[],"snapshot":{"statement":"A health technology company's quality assurance team builds and tears down test environments every day. Engineers currently create each environment by clicking through portal settings, and the hand-built environments drift apart over time. The QA lead wants repeatable deployments with no manual configuration. Which manageability capability should the team use?","options":["Automatic alerts based on configured resource metrics","Automatic replacement of resources that begin to fail","Manual configuration through the web portal, performed by senior engineers","Deploying resources from a preconfigured template, removing manual configuration"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-benefits-use-cloud-services/5-manageability-cloud","referenceTitle":"Describe the benefits of manageability in the cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00061","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:bdab8b68f977c997ec90c0795184ae3428c5db2095e79254ca59484e6faa45ab","aliases":[],"snapshot":{"statement":"A home-improvement retailer moves its inventory workloads to Azure on virtual machines. Its own IT staff installs and patches the guest operating systems, defines the virtual network layout, and configures the databases. The provider supplies the compute hardware, the connection to the internet, and the security of the physical facility.\n  Which layers remain the cloud provider's responsibility in this arrangement?","options":["Installing, configuring, and patching the guest operating systems","Designing the virtual network layout and configuring the databases","The physical hardware, the internet connection, and physical security","Licensing and patching the middleware and the development tools"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/2-describe-infrastructure-service/","referenceTitle":"Describe infrastructure as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00062","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:bb653fac35799af605a74a4a68736752770f21bba23b4adaf7d88c3d0fc4ec01","aliases":[],"snapshot":{"statement":"A junior cloud analyst is writing study notes about the three cloud service models for her team's onboarding guide. She wants one sentence that correctly captures what infrastructure as a service (IaaS) offers.\n  Which statement should she write?","options":["On-demand access to compute, storage, and networking, with the customer managing operating systems, applications, and data","A fully developed application that the customer only configures and uses, while the provider runs everything else","A managed platform where the provider handles operating systems and middleware so developers can focus on code","On-demand access to compute, storage, and networking, with the provider managing operating systems and databases"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/2-describe-infrastructure-service/","referenceTitle":"Describe infrastructure as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00063","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e1da80eddd21236d7600109b42f6f630e8bdcfe88a28a6c28d8b671cc889d88f","aliases":[],"snapshot":{"statement":"A regional hospital runs its patient-scheduling workloads on Azure virtual machines under the IaaS model. During a compliance review, the auditors want to confirm which security-related tasks stay with the hospital rather than with the cloud provider.\n  Which set of tasks remains the hospital's responsibility?","options":["Patching guest operating systems, managing physical security at the datacenter, and configuring virtual networks","Patching guest operating systems, configuring virtual networks, and hardening the scheduling applications","Maintaining internet connectivity, patching guest operating systems, and managing storage hardware","Patching guest operating systems, updating middleware, and managing the physical server hardware"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/2-describe-infrastructure-service/","referenceTitle":"Describe infrastructure as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00064","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:501f4da805682ea303e128cf3e70d8540222698ad56a1c44d6ec0c7adf847d43","aliases":[],"snapshot":{"statement":"A freight company rehosted its route-planning application from its own datacenter onto Azure virtual machines. The operations director is drafting a memo about what this move to IaaS means for the team's day-to-day work.\n  Which statement best describes the trade-off the team accepted?","options":["It gained the most control over the stack of any model, while offloading operating system patching to the provider","It reduced operational work to the minimum possible, while still keeping control of the operating systems","It transferred application maintenance to the provider, while keeping control of the physical hardware","It gained the most control over the stack of any model, while taking on the largest share of operational work"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/2-describe-infrastructure-service/","referenceTitle":"Describe infrastructure as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00065","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:1805cfc7f12ccb0b942ad9500f08ecf4c448360249c559ac0a2d1da7b152f63d","aliases":[],"snapshot":{"statement":"An architecture firm has always bought and housed its own rendering servers, which sit idle between project deadlines. The partners want to stop owning physical hardware altogether, but they insist on keeping every decision about what runs on it: which operating systems, drivers, and rendering software to use, and when to update them.\n  What does the firm get under the infrastructure as a service (IaaS) model?","options":["The provider owns the render hardware and also installs, patches, and updates the operating systems and rendering software on the firm's behalf","The firm keeps buying its render servers up front, and the provider merely houses and guards them between projects in its datacenter","The provider owns and maintains the hardware; what the firm runs on it — operating systems, drivers, and software — is up to the firm","The provider owns everything below the rendering application, and the firm only adjusts settings inside the application"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/2-describe-infrastructure-service/","referenceTitle":"Describe infrastructure as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00066","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:e9d10b6514baba96e6f4c4d471a83d20e5af21316a809a2d8a10550390a93382","aliases":[],"snapshot":{"statement":"An agritech startup is choosing how to host its new field-analytics web application. The option it prefers bundles the operating system, middleware, and development tools, all maintained by the provider; the startup would supply only its application code, data, and access configuration.\n  Which service model does this option represent?","options":["Infrastructure as a service (IaaS)","Platform as a service (PaaS)","Software as a service (SaaS)","A hybrid cloud deployment"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/3-describe-platform-service/","referenceTitle":"Describe platform as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00067","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:778ada3613f5a9030be41db07461f32c416e5c2ef891a976f4a761ebf84084d6","aliases":[],"snapshot":{"statement":"A smart-home device maker runs its product-portal application in a PaaS environment. A newly hired engineer asks who takes care of operating system and database patching now that the company no longer manages any servers.\n  What is the correct answer?","options":["The company patches them, because every layer above the hypervisor belongs to the customer","The provider patches the databases, while the company patches the operating systems","There is no operating system to patch; PaaS runs the application directly on the provider's hardware","The provider handles licensing and patching for operating systems and databases as part of the platform"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/3-describe-platform-service/","referenceTitle":"Describe platform as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00068","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:4a97ba07385405198eb2a0fe748291f493d241aeaa011ce1ff98f28dfa4e69bc","aliases":[],"snapshot":{"statement":"A streaming-media company builds its internal analytics dashboard on a PaaS platform. The engineering manager is writing role descriptions and wants to state what the team must concentrate on, as opposed to what the provider handles.\n  Which set of activities represents the customer's primary focus in PaaS?","options":["Writing application code, governing data, and managing access controls","Patching operating systems, maintaining middleware, and updating managed runtimes","Running the physical infrastructure, the internet connectivity, and the physical security","Writing application code, patching operating systems, and running the physical infrastructure"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/3-describe-platform-service/","referenceTitle":"Describe platform as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00069","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:bd44e191b3a842bcd1d259508e5a7891700f2d7fd0aed1182659fb725bc35de0","aliases":[],"snapshot":{"statement":"A software house is under pressure to shorten delivery time for a new customer portal. Its lead developer notes that on the PaaS platform under evaluation, scalability, high availability, and multitenant capability are provided as built-in features of the platform.\n  What does this mean for the development team's work?","options":["Developers write no application code at all; they only assemble built-in components into a product","Developers must still build scalability, high availability, and multitenancy into the application themselves","Developers write less code to deliver those capabilities, because the platform already supplies them","Developers must move the application onto virtual machines that they patch and maintain themselves"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/3-describe-platform-service/","referenceTitle":"Describe platform as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00070","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:450dcfd378ad93d56f97db9e23ae66d6d3109dfea0b1446cd19e319de2d0e710","aliases":[],"snapshot":{"statement":"An online marketplace currently hosts its storefront on Azure virtual machines, and its own staff patches the operating systems and maintains the runtime. The CTO plans to rehost the storefront on a PaaS offering instead, without changing the application itself.\n  Which responsibility moves to the provider as a result of this change?","options":["Protecting the application's data and managing access to it","Patching operating systems and maintaining the middleware and runtime","Writing, testing, and updating the application code","Paying for the resources that the workload consumes"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/3-describe-platform-service/","referenceTitle":"Describe platform as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00071","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6533eacfe92a5260fbffa84cade5483aacfa955ab4f44f2941396feec773b3dd","aliases":[],"snapshot":{"statement":"A dental clinic with no technical staff needs appointment-scheduling and patient-communication software. A vendor offers a complete application delivered over the internet: the clinic just creates user accounts, adjusts settings, and starts using it, while the vendor runs everything else.\n  Which service model is the clinic adopting?","options":["Software as a service (SaaS)","Infrastructure as a service (IaaS)","Platform as a service (PaaS)","Serverless computing (event-driven execution)"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/4-describe-software-service/","referenceTitle":"Describe software as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00072","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:a303ddbe27bf2cc1438a83602936a54c43a14e2c042cdb5de1f2cdd115ec1021","aliases":[],"snapshot":{"statement":"A small law office must deploy a client-document management tool quickly. It has no IT staff, and the partners have agreed to work with the vendor's standard feature set rather than request changes to the product.\n  Which statement about the software as a service (SaaS) model matches this situation?","options":["SaaS is the most flexible of the three models, so the office can modify the underlying platform","SaaS demands the most technical expertise of the three models to operate","SaaS is the least flexible model, but it is the easiest to get up and running","SaaS requires the customer to develop the application on a provider-managed framework"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/4-describe-software-service/","referenceTitle":"Describe software as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00073","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3fd1f2f7b2fa7e65960fe68eefb918db924ac8333cb8feac203b9ed9c1df261b","aliases":[],"snapshot":{"statement":"A university has adopted a SaaS collaboration platform for its administrative departments. The IT director is updating the operating handbook and wants to list what the university still needs to manage now that the platform is run by the vendor.\n  Which tasks remain primarily with the university?","options":["Infrastructure patching, application updates, and overall service availability","Data governance, identity and access settings, and device posture","Operating system licensing, database patching, and ongoing server maintenance","Data governance, operating system patching, and physical infrastructure"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/4-describe-software-service/","referenceTitle":"Describe software as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00074","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ad04c8ac4c1697893de55b06da3179be2a9ee286e041ff25fc4cfa356dd011cd","aliases":[],"snapshot":{"statement":"A hotel group is evaluating a SaaS property-management application. During the demo, the operations manager notices that the product supports only limited configuration options and cannot be reshaped around the group's older internal processes.\n  Which trade-off should the manager weigh before adopting it?","options":["The most management effort of the three models, in exchange for the most control over the platform","The most flexibility of the three models, in exchange for the customer patching the operating systems","No management effort at all, since even data protection passes to the provider","The least management effort of the three models, in exchange for the least flexibility among them"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/4-describe-software-service/","referenceTitle":"Describe software as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00075","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:33b5be489ba20d3b0db24c6546dfa10cfd9716206f7f387fe063bc2a196d6391","aliases":[],"snapshot":{"statement":"A municipal government is rolling out a SaaS records-management application across its departments. The records director asks how the application will be kept current once it is live, given that no one at the city installs or maintains any software.\n  What is accurate about how the application is kept current?","options":["The vendor keeps it current: feature updates and security improvements are applied automatically, and departments always use the latest version","The city's records team must schedule and install each feature update and every security fix during its own maintenance windows","The vendor maintains the infrastructure underneath, while the city's records team updates the application's features itself","The application stays at the version first deployed, so later security fixes become the city's own projects"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/4-describe-software-service/","referenceTitle":"Describe software as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00076","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:1a81bedf74927e501304a9e6a68de034e84e83284f416228174a9c57726ef293","aliases":[],"snapshot":{"statement":"A furniture manufacturer wants to move its legacy production-planning application to Azure. The application must run unchanged, on the same operating-system configuration it uses today, and the infrastructure team needs to keep managing everything from the operating system upward.\n  Which cloud service type should the manufacturer choose?","options":["Platform as a service (PaaS)","Software as a service (SaaS)","Serverless computing (event-driven execution)","Infrastructure as a service (IaaS)"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/2-describe-infrastructure-service/","referenceTitle":"Describe infrastructure as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00077","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a68c9b22c9e6b9f2f7f4dd3be6511e02f1721121b6a8613e79abb963d533088f","aliases":[],"snapshot":{"statement":"A fintech startup is building a new public web application. Its four-person engineering team wants a complete development environment in the cloud: they deploy their code and data while the provider maintains the operating systems, middleware, and development tools. No one can be spared for infrastructure work.\n  Which cloud service type fits this requirement?","options":["Infrastructure as a service (IaaS)","A hybrid cloud deployment","Software as a service (SaaS)","Platform as a service (PaaS)"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/3-describe-platform-service/","referenceTitle":"Describe platform as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00078","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:b2d7cc568f0c7d0d0a52278c7987a4852e06bf5d0d521c36d792e8090a8ee007","aliases":[],"snapshot":{"statement":"A 40-person insurance brokerage wants corporate email, calendars, and file sharing for everyone. It has no developers and no infrastructure team; the office manager just wants to create user accounts, set sharing policies, and have the service working the same week.\n  Which cloud service type fits best?","options":["Infrastructure as a service (IaaS)","Platform as a service (PaaS)","Software as a service (SaaS)","A private cloud deployment"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/4-describe-software-service/","referenceTitle":"Describe software as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00079","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:4349737e59026c0b19150eed54e22c511bea726e8359262e52613313c805b863","aliases":[],"snapshot":{"statement":"A game studio needs short-lived test environments that exactly replicate an established production configuration: a specific operating-system image, custom drivers, and pinned software versions. Environments are created and torn down on demand, and the QA team requires complete control over what runs inside them.\n  Which cloud service type meets this requirement?","options":["Infrastructure as a service (IaaS)","Platform as a service (PaaS)","Software as a service (SaaS)","Serverless computing (event-driven execution)"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/2-describe-infrastructure-service/","referenceTitle":"Describe infrastructure as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00080","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:53b03a6ef20dbbce7a453bbb3de7a43f9e59e5151140b0bd5d9b2d81cfc14824","aliases":[],"snapshot":{"statement":"A utility company's data team wants to analyze and mine its operational data, find patterns, and predict demand to improve planning. Rather than building and maintaining the underlying infrastructure, the team wants to develop its analysis solutions on a managed platform that already provides analytics services and tools.\n  Which cloud service type fits this scenario?","options":["Infrastructure as a service (IaaS)","Software as a service (SaaS)","A hybrid cloud deployment","Platform as a service (PaaS)"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/3-describe-platform-service/","referenceTitle":"Describe platform as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00081","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:640cb945daeb75aa0e00e46845da9ef982aa0b187f74c486cb68d640e127f203","aliases":[],"snapshot":{"statement":"A bank must host a proprietary risk engine whose vendor certifies it for exactly one legacy operating-system version, tuned with particular kernel settings. The certification was issued after an audit of that configuration, and it remains valid only while the bank's own platform team installs, configures, and tunes that operating system — if the provider managed it, the certified setup would be invalidated. The engine must keep running in Azure without being rewritten.\n  Which cloud service type can meet this requirement?","options":["Platform as a service (PaaS)","Infrastructure as a service (IaaS)","Software as a service (SaaS)","A hybrid cloud deployment"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/2-describe-infrastructure-service/","referenceTitle":"Describe infrastructure as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00082","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:467659ed3ace3bf67d59ef3c7039edff56be25559883ba1a77e39a54a75efe95","aliases":[],"snapshot":{"statement":"A news media company is launching a mobile app whose back end is a set of REST APIs. The engineering team wants to spend its time on the API code while the provider handles operating system and runtime management, including patching.\n  Which cloud service type fits this requirement?","options":["Infrastructure as a service (IaaS)","Software as a service (SaaS)","Platform as a service (PaaS)","A private cloud deployment"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/app-service/overview","referenceTitle":"App Service overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00083","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:35515f835ccb9a44fb1fdb153c6bc4730fc72742bbaff59ec8c20e129d74ca9a","aliases":[],"snapshot":{"statement":"A construction firm's finance department wants expense tracking and invoice approval for its site managers. Its processes follow standard accounting practices, no one on staff writes code, and the department wants a working product after only configuring categories and approval rules.\n  Which cloud service type fits best?","options":["Software as a service (SaaS)","Platform as a service (PaaS)","Infrastructure as a service (IaaS)","Serverless computing (event-driven execution)"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cloud-service-types/4-describe-software-service/","referenceTitle":"Describe software as a service — Microsoft Learn"}}
{"id":"RAN-AZ900-00084","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:7201302aa700c1075e18bc0849ba4d20ac2ca7cfe673acc784aa8d91d669fd74","aliases":[],"snapshot":{"statement":"A real-estate network has run its customer-relationship management (CRM) suite in its own datacenter for years; its small IT team spends much of each month patching the application, its databases, and the servers underneath. The board has approved replacing the suite with a vendor-delivered product: the vendor will host, run, update, and secure the application and everything beneath it, and the network's staff will only configure the product, manage user accounts, and govern its customer data. No one at the company will build or maintain any software for it.\n  Which cloud service type should the network adopt for the CRM workload?","options":["Infrastructure as a service (IaaS)","Software as a service (SaaS)","A hybrid cloud deployment","Platform as a service (PaaS)"],"answer":"B","reference":"https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-are-iaas-paas-and-saas","referenceTitle":"What are IaaS, PaaS, and SaaS? — Microsoft Azure"}}
{"id":"RAN-AZ900-00085","exam":"AZ900","topics":["cloud-concepts"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:c5fa60ca2fe3db122aa55a7dcf40bb8a850f94da8c1d8af009fc44892cbc8ad9","aliases":[],"snapshot":{"statement":"A grocery chain's cloud architect presents a migration plan: rehost the legacy warehouse system unchanged, with the team keeping control of its operating system; build the new customer portal so developers deliver code while the provider maintains the platform; and give HR a ready-made payroll and expense product.\n  Which assignment of cloud service types matches this plan?","options":["Warehouse: SaaS; portal: IaaS; HR: PaaS","Warehouse: IaaS; portal: PaaS; HR: SaaS","Warehouse: PaaS; portal: SaaS; HR: IaaS","Warehouse: IaaS; portal: SaaS; HR: PaaS"],"answer":"B","reference":"https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-are-iaas-paas-and-saas","referenceTitle":"What are IaaS, PaaS, and SaaS? — Microsoft Azure"}}
{"id":"RAN-AZ900-00086","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:31e9db75dcf4ca5c31131d5ae868b134de8cdef55af4f871937dd79a2d0188b9","aliases":[],"snapshot":{"statement":"A health-tech startup is preparing its first production deployment on Azure. During planning, the team lead asks what Azure means when documentation says that resources are deployed to a region such as Sweden Central.\n\nWhich description should the team lead use?","options":["A single physical Microsoft datacenter building that serves all customers in that part of the world","A collection of several Azure regions grouped together under one shared compliance boundary","A geographical area with one or more nearby datacenters, connected through a low-latency network","An isolated instance of Azure that is operated separately from the main public cloud"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/5-describe-azure-physical-infrastructure","referenceTitle":"Describe Azure physical infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00087","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:e696aaff3cce21d7b4cb866003d04ced34dad8f331ad6f9cdb5c412209f97e03","aliases":[],"snapshot":{"statement":"An online retailer runs its storefront in a single Azure region. The solutions architect is designing disaster recovery and wants to base it on the second region that Microsoft explicitly associates with the primary region for cross-region resilience.\n\nWhich Azure construct is the architect referring to?","options":["Region pairs, where each region is matched with another in the same geography for geo-replication and disaster recovery","Availability zones, which are groups of datacenters within a region that automatically replicate every resource to a second Azure region","Sovereign regions, which keep a private copy of a company's resources available for recovery purposes","Geographies, which automatically mirror all of a region's resources to a second geography for redundancy"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/reliability/regions-paired","referenceTitle":"Azure region pairs and nonpaired regions — Microsoft Learn"}}
{"id":"RAN-AZ900-00088","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d58f2e5c6ca14519c0145ef6f7b139f3877f0ccd102cb1f33b27519461c29185","aliases":[],"snapshot":{"statement":"A fintech's transaction platform runs in an Azure region that is paired. During an onboarding session, a new compliance officer asks what Microsoft's region-pairing design provides.\n\nWhich statement is accurate?","options":["Only one region in a pair can host customer workloads, because Microsoft reserves the second region for failover","Azure staggers planned updates across a region pair, rolling them out to one region at a time to reduce downtime risk","Deploying resources to a region in a pair automatically replicates them to the paired region and provides automatic failover","Paired regions are located in different geographies, so replication crosses compliance borders by design"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/reliability/regions-paired","referenceTitle":"Azure region pairs and nonpaired regions — Microsoft Learn"}}
{"id":"RAN-AZ900-00089","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d9aac8208a4d69976405acf44ba7a2dc98ad3d7778e207c15b36485f772c9028","aliases":[],"snapshot":{"statement":"A defense contractor delivers systems that process US federal government data. The contract requires an Azure environment that is physically and logically isolated from the public Azure cloud, with datacenters in the United States operated for government agencies and their partners.\n\nWhich option meets this requirement?","options":["Deploy to the West US region of the global Azure cloud and restrict access with strict network security group rules","Deploy to Azure in China, the sovereign instance that serves the Chinese market","Create a region pair by selecting two regions that are both located inside the United States","Use Azure Government, a dedicated cloud with physically isolated US datacenters and networks for government agencies and partners"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/azure-government/documentation-government-welcome","referenceTitle":"What is Azure Government? — Microsoft Learn"}}
{"id":"RAN-AZ900-00090","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d50d62b205e0e2bb206892124105b026d6054ded2801b188d8eab821dd68732e","aliases":[],"snapshot":{"statement":"A European media company is expanding its streaming service into mainland China. Its legal team reports that Azure services in China are delivered through a cloud instance that is physically separated from the global Azure cloud and independently operated by a local company.\n\nWhat is this instance called?","options":["Azure in China, a physically separated instance of cloud services independently operated and transacted by 21Vianet","The China East region of the global Azure public cloud, operated by Microsoft","Azure Government, the sovereign instance that Microsoft operates for government markets outside the United States","A region pair linking a Chinese region with a European region of the global cloud"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/china/overview-operations","referenceTitle":"Microsoft Azure operated by 21Vianet — Microsoft Learn"}}
{"id":"RAN-AZ900-00091","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:dc27bb4276fbfd35faa0a3e774f7b2ad8f127d0faf1b9e6c80c4b571cdba8a57","aliases":[],"snapshot":{"statement":"A logistics SaaS provider is designing a new deployment and selects a smaller Azure region close to its customers. While validating the design, the team finds that a specific virtual machine size it needs is not offered in that region, and it also plans to use Azure DNS.\n\nWhich explanation matches how Azure regions work?","options":["Every generally available Azure service and VM size is offered in every region; regional differences affect only pricing","Azure automatically deploys regional services into the nearest region that supports them, so the design needs no changes","Services and VM features vary by region and must be verified; global services like Azure DNS need no region selection","Missing services can be unlocked in any region by pairing it with a larger region in the same geography"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/5-describe-azure-physical-infrastructure","referenceTitle":"Describe Azure physical infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00092","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:93e05465d17c68475f6420e28d8f3f3db537409b6ff8232fd59c5821038d8a72","aliases":[],"snapshot":{"statement":"During onboarding at a university's IT department, a new cloud engineer joins the team's weekly Azure study group. In the first session, she is asked to define availability zones.\n\nWhich definition is correct?","options":["A separate Azure region, far from the primary one, that mirrors the region's resources for disaster recovery purposes","One or more physically separate datacenters within an Azure region, with independent power, cooling, and networking","A logical grouping of resources that share a lifecycle and are deployed and deleted together","A subscription-level policy that keeps all of a company's resources inside a single datacenter"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/5-describe-azure-physical-infrastructure","referenceTitle":"Describe Azure physical infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00093","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3ef0de0065301cac5aaa590cb0e763a5ba23518306466bbc771974c44d2a0b1b","aliases":[],"snapshot":{"statement":"An e-commerce company's checkout API runs in one Azure region. Engineering must keep checkout available if a single datacenter in that region fails, and all components must stay in the same region for low latency.\n\nWhat should the team do?","options":["Spread the application's components across availability zones within the region, replicating between zones","Deploy a complete second copy of the application to the region's paired region and fail over on every incident","Ask Azure support to pin all of the virtual machines to the region's most reliable datacenter","Place every virtual machine in a single availability zone to minimize inter-zone latency and rely on automatic recovery"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/reliability/availability-zones-overview","referenceTitle":"What are availability zones? — Microsoft Learn"}}
{"id":"RAN-AZ900-00094","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c63901537e6969928d9e6e1b64aecd17648e5b5b28851ddd53d92f2ef861a7a4","aliases":[],"snapshot":{"statement":"A solutions architect at an energy utility is finalizing a design for a region that supports availability zones and wants to confirm what Azure provides in such regions.\n\nWhich statement is accurate?","options":["The region provides exactly two zones: one that is active and one that is held on standby","The zones share power and network infrastructure to reduce cost, so they recover as a unit","Each zone is itself a separate Azure region, and the set of zones together forms a region pair","The region provides at least three zones, each with independent power, cooling, and networking"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/5-describe-azure-physical-infrastructure","referenceTitle":"Describe Azure physical infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00095","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:fbe23209224211490bc0209b36283fce789349260d5d973ef9671ba8b3cc005c","aliases":[],"snapshot":{"statement":"After a storm briefly disrupted one Azure region, an insurance company's review board asks its cloud team to explain the difference between availability zones and region pairs in terms of the failures they address.\n\nWhich explanation is correct?","options":["Availability zones protect against full-region outages, so replicating to a paired region adds little value","Availability zones address datacenter and zone failures within a region, while replicating to a second region addresses region-wide failures","Region pairs protect against datacenter-level failures, while availability zones exist to survive outages that affect an entire region","Neither mechanism helps with unplanned failures; both exist only to sequence planned maintenance"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/reliability/availability-zones-overview","referenceTitle":"What are availability zones? — Microsoft Learn"}}
{"id":"RAN-AZ900-00096","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:ea8ec5bab624d32160dcabd62fbba61793490703f00b857c015c470fa66296fe","aliases":[],"snapshot":{"statement":"An intern at a logistics company has just started learning about cloud infrastructure. At the end of her first week, she asks her mentor what an Azure datacenter actually is.\n\nWhich answer should the mentor give?","options":["A physical facility with servers arranged in racks and dedicated power, cooling, and networking infrastructure","A logical container that groups related Azure resources so they can be managed together with shared access control","A geographical area containing at least one, but potentially multiple, datacenters","A billing unit that aggregates the costs of all resources an organization deploys"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/5-describe-azure-physical-infrastructure","referenceTitle":"Describe Azure physical infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00097","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:1bdc15ed84f94e58d965b793dae22f32c268e7894c66d73a5d68ffb32e03cfff","aliases":[],"snapshot":{"statement":"A junior administrator at a marketing agency is about to deploy the company's first virtual machine and asks how to choose which specific Azure datacenter will host it.\n\nWhat should the team lead tell her?","options":["Open a support ticket so that Microsoft can assign a dedicated datacenter building to host your company's subscription resources","Create a separate resource group for each datacenter, since resource groups map one-to-one to datacenters","You don't select datacenters individually — you choose a region, and where needed an availability zone; Azure manages placement","Choose the management group that is hosted in the datacenter closest to the agency's office"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/5-describe-azure-physical-infrastructure","referenceTitle":"Describe Azure physical infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00098","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6ec3e39e2cff4cfbeffd148e511b62b284cf3ba1249a081b187d086e4acd5195","aliases":[],"snapshot":{"statement":"A video-streaming startup is launching in South America, with nearly all viewers in Brazil. The team wants to minimize the network distance between its viewers and the virtual machines that serve the video segments.\n\nHow does the region choice affect where the workloads physically run?","options":["Azure distributes the resources evenly across all regions in the Americas to balance global load","The resources run in datacenters inside the chosen region, such as Brazil South, physically close to the viewers","The resources run in the nearest sovereign region, because Brazil South is reserved for government workloads","The physical location is irrelevant, because Azure always runs resources in the region where the subscription was created"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/5-describe-azure-physical-infrastructure","referenceTitle":"Describe Azure physical infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00099","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:0fe6240dc4879c36ae4eed2e44c2e8b667cc16f2a892a3a9a1c3600c9db89716","aliases":[],"snapshot":{"statement":"During onboarding at a manufacturing company, new engineers take a short Azure terminology quiz as part of their induction. One of the quiz questions asks what counts as an Azure resource.\n\nWhich choice is correct?","options":["Only the physical hardware, racks, and facilities that Microsoft operates inside its datacenters","Only compute services such as virtual machines; storage and networking are managed by Microsoft and aren't resources","The subscriptions and billing accounts that are used to group an organization's workloads together","Manageable items that you create and deploy through Azure, such as virtual machines, storage accounts, web apps, and databases"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/overview","referenceTitle":"What is Azure Resource Manager? — Microsoft Learn"}}
{"id":"RAN-AZ900-00100","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:620f7813df35d1152bf40526fb127306820e416afd8c1f807725534263322138","aliases":[],"snapshot":{"statement":"A junior administrator at a hospital trust is documenting how resource groups work. Her draft says that a virtual machine can belong to two resource groups at once, that each group can hold only one resource type, and that groups can be nested inside each other.\n\nWhich correction should her reviewer make?","options":["Each resource is in exactly one resource group at a time; groups can hold different resource types and can't be nested","A resource can belong to several resource groups at the same time, as long as all of those groups are located in the same region","Each resource group is limited to one resource type, but nested groups can be used to model departments","Resource groups are optional, and any resource can be deployed directly under a subscription without a group"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/6-describe-azure-management-infrastructure","referenceTitle":"Describe Azure management infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00101","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:80e20570660b9c982963f2025ae4e5e5c44dc1e5c1b652522f1b661c2899fbba","aliases":[],"snapshot":{"statement":"A software company reorganizes its teams, and the resources of a transferred project — virtual machines, storage accounts, and a database — must now live in a different resource group. All of the involved resource types support the move operation, and the team wants to keep the same resources running, without recreating anything.\n\nWhat is the correct approach?","options":["Export the resources' configurations, delete the resources, and redeploy everything inside the new group","Keep the resources where they are, because resources stay in the group where they were created for their entire lifetime","Use the Azure move operation to transfer the resources into the new resource group, keeping the same resources running","Move the whole subscription into a management group owned by the receiving team, which transfers the resource group along with it"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/move-resource-group-and-subscription","referenceTitle":"Move Azure resources to a new resource group or subscription — Microsoft Learn"}}
{"id":"RAN-AZ900-00102","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:c0f8c78fcfb0941434c3ba6e99b08d03963a96e1235b2ce72fd3a8d66dde39d0","aliases":[],"snapshot":{"statement":"A payments team at a fintech wants to keep a virtual machine in West Europe and a storage account in North Europe inside one resource group named rg-payments, so the team can manage access in one place and delete everything together when the project ends.\n\nWhich assessment of this plan is accurate?","options":["The plan fails, because each resource group is created in one region, and every resource inside it must be in that same region","The plan works: a resource group can hold resources in different regions, and deleting the group deletes the resources inside it","The plan works only if the two resources are first replicated into the same availability zone","The plan works, but deleting the group only removes the group's metadata and leaves both resources running"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/overview","referenceTitle":"What is Azure Resource Manager? — Microsoft Learn"}}
{"id":"RAN-AZ900-00103","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:513704a3fa534f2120f44faeb1228b64389928fcd966627c551294e0057b7347","aliases":[],"snapshot":{"statement":"The CFO of a mid-sized energy company is new to Azure and asks what the company actually receives when it signs up for an Azure subscription.\n\nWhich explanation is correct?","options":["A unit of management, billing, and scale, linked to an Azure account, that provides access to Azure services","A logical container that groups related resources so that they can be deployed, updated, and deleted together","A governance scope that sits above subscriptions to enforce policies across many of them","A physical grouping of datacenters connected by a low-latency network within a geography"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/6-describe-azure-management-infrastructure","referenceTitle":"Describe Azure management infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00104","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:90b8a6058b466888cf41e05594faf96b761d9607d3431dff957c8d1a13747b92","aliases":[],"snapshot":{"statement":"A SaaS company wants its production platform and its experimental sandbox to be billed as completely separate units, each with its own invoice and its own access-control rules.\n\nWhich approach should the platform lead take to meet both requirements with the least administrative effort?","options":["Create one resource group per environment and ask Azure support to issue invoices for each resource group","Create one management group per environment, and place all of the environment's resources directly inside the management group","Create a separate subscription for each environment, giving each its own billing reports and invoices and its own access rules","Register a second Azure account in a brand-new tenant and rebuild the sandbox environment there"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/6-describe-azure-management-infrastructure","referenceTitle":"Describe Azure management infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00105","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e2872b97e0624c69105c6b8196231a260063a30655d256d52e1d1b7437ead105","aliases":[],"snapshot":{"statement":"An administrator at a manufacturing company moves a virtual machine from the resource group rg-prototype to the resource group rg-factory; both groups are in the same Azure subscription. A finance analyst asks how the move affects billing.\n\nWhich response is correct?","options":["Each resource group generates its own separate invoice, so all future charges for the virtual machine will appear on a new bill","The charges are recalculated using the pricing of the region where the new resource group is located","The charges now roll up to the management group, which consolidates one invoice for all subscriptions beneath it","The charges still roll up to the same subscription, because subscriptions — not resource groups — are the billing boundary"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/6-describe-azure-management-infrastructure","referenceTitle":"Describe Azure management infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00106","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:081b27198170b05e2f3c9f9b4278014c817b0bc74751909633728f02d1a15a46","aliases":[],"snapshot":{"statement":"An energy conglomerate has grown to dozens of Azure subscriptions across its business units. The cloud director asks what Azure management groups would give the company.\n\nWhich description is correct?","options":["Logical containers that group related resources so they can be deployed and deleted together","Governance scopes above subscriptions that manage access, policies, and compliance by inheritance","Paired groups of datacenters that provide disaster recovery for the data of every subscription","Billing plans that merge the invoices of all the company's subscriptions into a single enterprise statement"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/governance/management-groups/overview","referenceTitle":"What are Azure management groups? — Microsoft Learn"}}
{"id":"RAN-AZ900-00107","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ba023a752e11bca48e182a2e3c77d4e1639185ca1e4801a103a10a96a8df3fa6","aliases":[],"snapshot":{"statement":"A bank's security team must enforce an Azure Policy that restricts which regions can be used for new resources. All of the bank's subscriptions are organized under a single management group, and the policy must cover every subscription beneath it — including subscriptions added there later — with a single assignment.\n\nWhere should the team assign the policy?","options":["At each subscription individually, and repeat the assignment manually whenever a new subscription is created","At each resource group, because Azure Policy can't be applied above the resource group scope","At the region level, so the policy applies to every resource deployed in the regions the bank uses","At the management group level, so every nested subscription, resource group, and resource inherits it, including subscriptions added later"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/governance/management-groups/overview","referenceTitle":"What are Azure management groups? — Microsoft Learn"}}
{"id":"RAN-AZ900-00108","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:4795f160e224fc19f1c220b7144f0183eebf9c63712cc55a1d2d94ca7de85817","aliases":[],"snapshot":{"statement":"An enterprise architect at a multinational retailer is reviewing a colleague's notes about management groups before the company's governance design is finalized. The notes must be corrected wherever they are wrong.\n\nWhich note is accurate?","options":["Management group hierarchies have unlimited depth, and a single subscription can report to several management groups at the same time","Management groups can contain resource groups and individual resources directly, alongside subscriptions","A management group tree can be six levels deep (excluding root and subscription levels), and each subscription has only one parent","Each Microsoft Entra tenant supports several independent management group hierarchies, one per region"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/governance/management-groups/overview","referenceTitle":"What are Azure management groups? — Microsoft Learn"}}
{"id":"RAN-AZ900-00109","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7ea443249542a9f2a62075cfd0cc1ea73b26f6fb9c000539a2ede60de8f0bc8a","aliases":[],"snapshot":{"statement":"On the first day at a telecom company, a new cloud engineer sketches Azure's management hierarchy on a whiteboard but is unsure about the order of the levels.\n\nWhich ordering nests the scopes correctly, from highest to lowest?","options":["Subscriptions → management groups → resource groups → resources","Management groups → resource groups → subscriptions → resources","Resource groups → subscriptions → management groups → resources","Management groups → subscriptions → resource groups → resources"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-core-architectural-components-of-azure/6-describe-azure-management-infrastructure","referenceTitle":"Describe Azure management infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00110","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:cb678e74a5aad303695eab6e78e0e811c417be41b76209f58be20d6f91c0b550","aliases":[],"snapshot":{"statement":"An auditor at a pharmaceutical company is reviewing a storage account. The company assigned the compliance team's Reader role at the management group level and assigned an allowed-regions policy at the subscription level. The auditor wants to know which of these assignments affect the storage account.\n\nWhich statement is correct?","options":["Both affect it: role assignments and policies from higher scopes are inherited by the resource groups and resources beneath them","Only the policy applies, because role assignments made at a management group level never reach down to individual resources","Only the role assignment applies, because policies are enforced solely at the scope where they were assigned","Neither applies until the storage account's resource group is explicitly added to both assignments"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/overview","referenceTitle":"What is Azure Resource Manager? — Microsoft Learn"}}
{"id":"RAN-AZ900-00111","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:2d14e50ce4f9b0258377354909e584298599fe0719877e369aba47aabe74ecbd","aliases":[],"snapshot":{"statement":"A logistics company is choosing a compute option for a new shipment-tracking service. The service is packaged as a self-contained unit that bundles the application code and its dependencies, several instances of it will run on the same host machine, and the operations team does not want to install, patch, or manage an operating system for each running instance. Which compute type meets these requirements?","options":["Virtual machines, because each instance runs its own operating system that the team can connect to and manage","Serverless functions, because the code runs only when events arrive and no infrastructure is provisioned while idle","Containers, because they are lightweight environments that scale out and stop dynamically with no operating system to manage per instance","Azure Virtual Desktop, because it delivers full Windows desktops and applications to users from the cloud"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/5-containers/","referenceTitle":"Describe Azure containers — Microsoft Learn"}}
{"id":"RAN-AZ900-00112","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0b19d213a91dd01783ed73b454439538f0d02ae2e195b94a0381fb40a67b3085","aliases":[],"snapshot":{"statement":"An insurance company receives claim documents at unpredictable times throughout the day. Each new document must be validated and converted a few seconds after it lands in a storage container. The company wants to avoid paying for compute that sits provisioned but idle between documents. Which Azure compute service should the company use?","options":["A virtual machine that runs a validation service continuously and polls the storage container for new documents","An Azure function that wakes up when a new document arrives, with no resources kept provisioned while no events occur","A virtual machine scale set that keeps a minimum number of instances running at all times to guarantee instant pickup","An always-on container group that listens for new documents and processes them as they appear"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/6-functions/","referenceTitle":"Describe Azure functions — Microsoft Learn"}}
{"id":"RAN-AZ900-00113","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:74c6ff4d30443e70763f9adf97d93da01d8c446cf8215ebc550302f801529ad2","aliases":[],"snapshot":{"statement":"A manufacturer is moving a simulation application to Azure. The application is certified for a specific Linux distribution, and its installation procedure requires the team to tune kernel parameters, install a custom device driver, and sign in to the server's operating system directly to troubleshoot. The team is deciding between virtual machines and containers. Which option should they choose, and why?","options":["Virtual machines, because each one runs its own operating system, which the team can connect to and manage","Containers, because they offer the same operating system access as virtual machines, only with lower overhead","Containers, because an application certified for a specific Linux distribution runs unchanged inside any container image","Virtual machines, because Linux-based workloads cannot be packaged as container images in Azure"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/5-containers/","referenceTitle":"Describe Azure containers — Microsoft Learn"}}
{"id":"RAN-AZ900-00114","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e1f2a69fc9c619b3cbe62b4f58af384da9916f9e4943696dce730cd3a6c49589","aliases":[],"snapshot":{"statement":"A gaming studio is reviewing which of its workloads would fit Azure Functions, the serverless compute option. Three of the following workloads are good candidates for a function triggered by an event; one is a poor fit. Which workload is the poor fit?","options":["Generate a thumbnail image a few seconds after a player uploads a new profile picture","Send a confirmation message to the customer whenever a new order arrives in the online store","Run a database cleanup routine once per night, started on a schedule","Host a game-server process that must run around the clock, holding match state in memory for hours"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/6-functions/","referenceTitle":"Describe Azure functions — Microsoft Learn"}}
{"id":"RAN-AZ900-00115","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:9e5acc762ccf746aadfa4f7b77f50cab26e72462ade87c6dfe06a3fb75781baf","aliases":[],"snapshot":{"statement":"A cloud architecture intern is writing a one-page summary of the differences between virtual machines, containers, and Azure Functions for the team wiki. Which summary is accurate?","options":["Virtual machines abstract the operating system away from the customer, while containers require managing a guest operating system for each instance","Virtual machines provide the most control, containers remove the per-instance operating system, and functions run code in response to events","Containers provide the most control over the computing environment, because each container ships with a dedicated operating system for one application","Functions keep their resources provisioned at all times, which makes them the natural choice for applications that must run continuously without breaks"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/5-containers/","referenceTitle":"Describe Azure containers — Microsoft Learn"}}
{"id":"RAN-AZ900-00116","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:cb1c1db31eab48f08a163f460dec66aef49438d7cf88e9714d56943fd374ebc6","aliases":[],"snapshot":{"statement":"A retailer wants to move an existing Windows server application to Azure with minimal changes. The application is installed directly on the server's Windows operating system, and the IT team plans to keep applying operating system patches and running their existing OS-level monitoring tools, exactly as they do in their own datacenter today. Which compute type supports this approach?","options":["Azure Functions, because the server application's logic can be split into small event-driven pieces","Azure Virtual Desktop, because the application is delivered to users as a published remote desktop","Virtual machines, because this infrastructure as a service option leaves the operating system to the customer","Azure App Service, because it runs Windows-based applications without the customer managing any infrastructure"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/2-virtual-machines/","referenceTitle":"Describe Azure virtual machines — Microsoft Learn"}}
{"id":"RAN-AZ900-00117","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:85adfe5f9bb90355b777ccb0eeb30d5ade69d2f4cfb0cd62fec9a9574cf136d2","aliases":[],"snapshot":{"statement":"A training provider is comparing Azure compute services for a student enrollment portal. The portal needs a service that creates and manages a group of load balanced virtual machine instances, where the number of instances automatically increases or decreases in response to demand or a defined schedule. Which Azure service provides this behavior?","options":["Azure Virtual Machine Scale Sets, for groups of VM instances","Azure Virtual Desktop, for delivering cloud-hosted Windows desktops and apps to users","Availability sets, for protecting VMs from datacenter failures and maintenance","Azure App Service, for managed hosting of web applications and APIs"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/virtual-machine-scale-sets/overview","referenceTitle":"Azure Virtual Machine Scale Sets overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00118","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:42c8b4c335c25f3ffc3332d474d25e075505cb74a7bfc989403067f46170cd5a","aliases":[],"snapshot":{"statement":"A media streaming startup runs its website front end on Azure virtual machines. Traffic is low overnight, spikes sharply during evening hours and live events, and manual resizing is error-prone. The team wants the platform to add virtual machine instances when demand rises and remove them when it falls, while distributing incoming requests across the instances. Which Azure option should the team use?","options":["Place the front-end virtual machines in an availability set, which adjusts the instance count automatically during peaks","Keep a fixed set of extra-large virtual machines running at all times, sized for the evening and event peaks","Deploy the front end as a Virtual Machine Scale Set, so load balanced instances automatically follow demand","Spread the front-end virtual machines across availability zones, so each zone adds instances during peak hours"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/virtual-machine-scale-sets/overview","referenceTitle":"Azure Virtual Machine Scale Sets overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00119","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:397bf24b5f892091195dd16b1bb59b5ece89bfcec08b95b9a1eb530313195b02","aliases":[],"snapshot":{"statement":"A regional bank runs a multi-tier application on several virtual machines in a region that does not support availability zones. Operations wants to ensure that planned maintenance never restarts all of the tier's virtual machines at the same time, and that a single rack power or network failure cannot affect every machine. The deployment must stay as it is: a small group of differently configured virtual machines that admins manage individually, with no automatic scaling. What should the team use?","options":["An availability zone, so each virtual machine runs in a physically separate location within the region","An availability set, which spreads the virtual machines across fault domains and update domains","A Virtual Machine Scale Set, which replaces the tier with instances that grow and shrink automatically","Azure Virtual Desktop, which centralizes the tier's virtual machines in a managed cloud environment"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/virtual-machines/availability-set-overview","referenceTitle":"Availability sets overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00120","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:1b7ac2999999dc8d684313010ca02a3b57cad0e0e807afa6e58664a540745277","aliases":[],"snapshot":{"statement":"An administrator is documenting how availability sets protect the company's virtual machines within a datacenter. When describing fault domains in the document, which explanation is correct?","options":["A fault domain is a group of virtual machines that are restarted at the same time during planned maintenance","A fault domain is a pair of Azure regions that host copies of each other's workloads for disaster recovery","A fault domain is a set of virtual machines that share the same size and must be upgraded together","A fault domain is a group of virtual machines that share a common power source and network switch"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/virtual-machines/availability-set-overview","referenceTitle":"Availability sets overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00121","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:00e2041ab2707a92fa4df24eb453e5db64bdd7a26d037265b0bbc10e5efa1f1f","aliases":[],"snapshot":{"statement":"A nonprofit's staff work from offices, home, and field sites on many different devices. The organization wants to give each user a full Windows desktop with the organization's applications, keep those desktops running in the cloud rather than on local devices, and let users connect from any device through an app or a web browser. Which Azure service should the nonprofit use?","options":["Azure Virtual Machine Scale Sets, for autoscaling groups of VM instances","Availability sets, for protecting VMs from hardware failures and maintenance","Azure Virtual Desktop, for cloud-hosted desktops and apps","Azure App Service, for hosting web applications and APIs"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/virtual-desktop/overview","referenceTitle":"What is Azure Virtual Desktop? — Microsoft Learn"}}
{"id":"RAN-AZ900-00122","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:938a3970bd68cb81b1e53932572706d9b5e5b757d8041a66b3041cd23e29f2ac","aliases":[],"snapshot":{"statement":"A junior architect is writing runbook notes about availability sets for the operations team. Which statement belongs in the runbook?","options":["An availability set limits the impact of hardware failures, power interruptions, and maintenance, but not operating system or application failures","An availability set replicates virtual machines to a paired region automatically, keeping a warm copy ready for failover","An availability set keeps the application online by automatically replacing any failed virtual machine with a new one in the set","An availability set protects against application crashes by restarting the crashed application on a healthy machine in the set"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/virtual-machines/availability-set-overview","referenceTitle":"Availability sets overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00123","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c1b40b36f50013e8a2c533d92cee326a185c38a6f015fcc35f605509350a4106","aliases":[],"snapshot":{"statement":"A hospital's clinical staff use shared workstations and need only two line-of-business applications during their shifts; they do not need a full remote desktop. The IT department wants to deliver just those applications from Azure so they appear on the staff's devices, while the underlying Windows environment stays centralized in the cloud. Which capability of Azure Virtual Desktop should the IT department use?","options":["Publish a full Windows desktop to every clinical user, and instruct staff to open the two applications inside it","Move the two applications into container images and run them in Azure Container Instances","Publish the two applications individually, so they run in the cloud and appear on staff devices without a full desktop","Rebuild the two applications as web APIs and host them in Azure App Service"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/virtual-desktop/overview","referenceTitle":"What is Azure Virtual Desktop? — Microsoft Learn"}}
{"id":"RAN-AZ900-00124","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:960fad347e1f47ec0cbb08f01aa76ca3c4b2b06393a5a95131217106f2724091","aliases":[],"snapshot":{"statement":"A team creating its first Azure virtual machine is deciding how to handle the machine's storage. They want disk volumes where the team only specifies the disk type and the disk size, and Azure handles creating and managing the underlying storage for them. Which Azure feature provides this experience?","options":["Unmanaged disks placed in a storage account that the team itself creates and maintains manually","The virtual machine's temporary disk, which Azure manages for storing long-term application data","Availability sets, which distribute the machine's disks across the datacenter automatically","Azure managed disks, block-level storage volumes that Azure creates and manages on the team's behalf"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/virtual-machines/managed-disks-overview","referenceTitle":"Overview of Azure Disk Storage — Microsoft Learn"}}
{"id":"RAN-AZ900-00125","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:2b1e9c2c39309a0ab624523ccb3e1953049a18f14eaf134680b35c261893df2c","aliases":[],"snapshot":{"statement":"An analytics company is sizing a virtual machine for an in-memory database that keeps large working sets in RAM and runs memory-intensive analytics queries. According to Azure virtual machine size families, which family should the team evaluate first for this workload?","options":["The F-series of compute-optimized sizes, designed for CPU-intensive application tiers","The E-series of memory-optimized sizes, designed for in-memory databases and analytics workloads","The B-series of burstable sizes, designed for dev and test workloads with occasional CPU spikes","The N-series of GPU-enabled sizes, designed for AI training, inference, and graphics workloads"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/2-virtual-machines/","referenceTitle":"Describe Azure virtual machines — Microsoft Learn"}}
{"id":"RAN-AZ900-00126","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f02187ea8daec731c0495b391da6798226d441aa2d3823071eeb6cd294763db5","aliases":[],"snapshot":{"statement":"A software company wants every new virtual machine in its test lab to start from the same baseline: a specific Windows Server version plus the company's monitoring agent and standard tools already installed. According to Azure virtual machine concepts, what should the team use so that new virtual machines are created from this starting point?","options":["An image, which is a template that already includes an operating system and tools","An availability set, which applies one standard configuration to every virtual machine added to it","A scale set, which clones the disk of an existing virtual machine to each new instance","A network security group, which enforces the same software settings on the machines in its scope"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/2-virtual-machines/","referenceTitle":"Describe Azure virtual machines — Microsoft Learn"}}
{"id":"RAN-AZ900-00127","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:81ec7609d12b4d93237cb8c2e625296ab0f13ea6c9152218afd60a3f2c1a2d25","aliases":[],"snapshot":{"statement":"A network engineer is reviewing the supporting resources that Azure creates alongside a new virtual machine so that the machine can communicate on a virtual network. Which resource connects the virtual machine to its virtual network?","options":["The public IP address, which links the virtual machine directly to the virtual network's address space","The network security group, which connects the virtual machine to its subnet and filters its traffic","The subnet, which plugs the virtual machine into the address space of the virtual network","The network interface card, which is the resource that connects the virtual machine to the virtual network"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/virtual-machines/overview","referenceTitle":"Overview of virtual machines in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00128","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:24cc1a2a6b4516450dfd2777f74f31f0fe92745cc4e1c026d227ebf25f5f450f","aliases":[],"snapshot":{"statement":"A startup is building a customer-facing web application with a RESTful API backend. The developers want to write code in the programming language of their choice and have Azure run the hosting infrastructure for them, with automatic scaling and high availability built into the platform. Which Azure service is designed for this?","options":["Azure Container Instances","Azure App Service","Azure Virtual Desktop","Azure Virtual Machine Scale Sets"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/7-describe-application-hosting-options/","referenceTitle":"Describe application hosting options — Microsoft Learn"}}
{"id":"RAN-AZ900-00129","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:61d26b97a3e3de026a25fbb66864fdc3e626e30d00e8549b5838af8dba00bfb5","aliases":[],"snapshot":{"statement":"An enterprise hosts a third-party Windows application on premises. The vendor's support contract requires full remote access to the server, including permission to install tools that modify the Windows registry and to restart the operating system during troubleshooting. When the company moves this application to Azure, which hosting option satisfies the support requirements?","options":["A virtual machine, because it gives maximum control of the hosting environment, including the guest operating system","Azure App Service, where the vendor's Windows application would run on shared infrastructure that Microsoft operates","Azure Container Instances, because containers isolate the application from the underlying operating system","Azure Functions, because serverless execution removes all operating system responsibilities from the vendor"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/7-describe-application-hosting-options/","referenceTitle":"Describe application hosting options — Microsoft Learn"}}
{"id":"RAN-AZ900-00130","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:bf4c4d955a55e34dd0c4e11749808404028a4800423d1b6c5c7f35a267907428","aliases":[],"snapshot":{"statement":"A data engineering team has packaged a one-off data transformation job as a single container image. They need to run it in Azure for a few hours tonight, without provisioning virtual machines, and without setting up a container orchestration platform they would then have to learn and operate. Which service should the team use?","options":["Azure Kubernetes Service (AKS), which manages the lifecycle of a fleet of containers","Azure Virtual Machine Scale Sets, which run the job across automatically scaled virtual machine instances","Azure App Service, which hosts the container behind a managed web endpoint","Azure Container Instances, which runs a single container without managing virtual machines or adopting a higher-level service"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/container-instances/container-instances-overview","referenceTitle":"Serverless containers in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00131","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:be3c9029b9e611b7af76e4c968803596c963ca9d98875fdfdfd768da99369d3e","aliases":[],"snapshot":{"statement":"A software company runs a microservices platform made of dozens of containerized services. Deployments require coordinated rollouts, automatic restarts of unhealthy containers, scaling of individual services, and routing of traffic between them. The company wants Kubernetes-style orchestration but does not want to run the Kubernetes control plane itself. Which Azure service should the company choose?","options":["Azure Container Instances, because it orchestrates fleets of containers with built-in load balancing","Azure Kubernetes Service (AKS), the managed Kubernetes service that runs the control plane for the company","Azure Virtual Machine Scale Sets, because they orchestrate container images as virtual machine instances","Azure App Service, because it orchestrates containerized web applications on managed infrastructure"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/aks/intro-kubernetes","referenceTitle":"What is Azure Kubernetes Service (AKS)? — Microsoft Learn"}}
{"id":"RAN-AZ900-00132","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:09c827cf946c031fd2d9aacfb2530fb392630599d10bbefcf6c26f44358b915a","aliases":[],"snapshot":{"statement":"A retailer's platform team is decomposing a monolith into containerized microservices. The team does not want to manage virtual machines, does not want to learn and operate a full container orchestration system, but still needs each microservice to scale with demand and spread requests across instances without building that plumbing itself. Which Azure service best matches these needs?","options":["Azure Kubernetes Service (AKS), whose full Kubernetes orchestration the team would have to learn and operate","Azure Virtual Machine Scale Sets, which host each microservice on automatically scaled virtual machine instances","Azure Container Apps, which includes built-in load balancing and scaling without the overhead of a full orchestration platform","Azure Container Instances, which runs single containers without built-in load balancing and scaling"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/5-containers/","referenceTitle":"Describe Azure containers — Microsoft Learn"}}
{"id":"RAN-AZ900-00133","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f419bd78dc21063c44bd9badddbabbc2a914b92988c29a189f05310762596ccc","aliases":[],"snapshot":{"statement":"A fintech is launching a public web portal with a REST API backend. The code is written in Java, the team stores its source in GitHub, and they want automated deployments, automatic scaling, and high availability without patching servers or container hosts. Which hosting approach best fits this launch?","options":["Azure App Service, which hosts web apps and APIs with automatic scaling, high availability, and automated deployment from GitHub","A virtual machine, which gives the team full control of the operating system and the hosting environment","Azure Container Instances, which runs each API endpoint as a separate container on demand","Azure Kubernetes Service (AKS), which orchestrates the portal and API as containerized workloads"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-compute-networking-services/7-describe-application-hosting-options/","referenceTitle":"Describe application hosting options — Microsoft Learn"}}
{"id":"RAN-AZ900-00134","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:7fe3bb3db34062f10282b945fe77159924d26a701cd63a7784ab8feab7be9a9f","aliases":[],"snapshot":{"statement":"A company is planning its first Azure deployment, which will include several virtual machines and a managed database. The cloud architect tells the team they must create an Azure Virtual Network before deploying these resources. What is the purpose of that virtual network?","options":["To reserve a range of public IP addresses that Azure assigns automatically to every resource created in the subscription","To give Azure resources like virtual machines a private network for secure communication with each other, the internet, and on-premises networks","To encrypt traffic between the company's branch offices and the internet by using the company's own VPN device","To resolve the domain names of the company's public website through Azure's global network of DNS name servers"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/virtual-network/virtual-networks-overview","referenceTitle":"What is Azure Virtual Network? — Microsoft Learn"}}
{"id":"RAN-AZ900-00135","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:af8b94787e75cf0b293a48e5898d17f7d6da3bd5967d742db21068b1cdfacbe2","aliases":[],"snapshot":{"statement":"A retailer has created an Azure Virtual Network with a defined IP address space for its online store. The team wants the web servers and the database servers in separate segments of that address space, so each group can be organized and secured as a unit while remaining in the same network. What should the retailer create inside the virtual network?","options":["A peering connection to a second virtual network that hosts only the database servers","An ExpressRoute circuit dedicated to the database servers and their clients","A VPN gateway placed in front of the database servers","Subnets, which divide the virtual network's address space into segments that resources are deployed into"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-networking-services/2-virtual-network/","referenceTitle":"Describe Azure virtual networking — Microsoft Learn"}}
{"id":"RAN-AZ900-00136","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:1c20e514e2d4caf61f3467935ba9201a8aa40bc031d5e8e83b2c155ec56c6486","aliases":[],"snapshot":{"statement":"An organization's Azure environment has two virtual networks: one for production workloads and one for shared services, deployed in different regions. Resources in the production network must communicate directly with resources in the shared services network as if they were on the same infrastructure. Which Azure networking feature connects the two virtual networks so that resources in each can communicate?","options":["Virtual network peering, which links the two networks directly","A network security group applied to both virtual networks","An ExpressRoute circuit connecting the two regions","An Azure DNS zone shared by both virtual networks"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/virtual-network/virtual-network-peering-overview","referenceTitle":"Azure Virtual Network Peering — Microsoft Learn"}}
{"id":"RAN-AZ900-00137","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:8fffb2a131a4256ceb878b06657083faa456021f364e9b3b2bf8b724ae34df6e","aliases":[],"snapshot":{"statement":"A multinational company has virtual networks in two Azure regions. Applications in each region must exchange data, and the security team requires that this traffic never enters the public internet; the team also does not want to deploy virtual network gateways for this purpose. Which approach meets the requirements?","options":["Connect the virtual networks with a virtual network-to-virtual network VPN gateway connection, which encrypts the traffic as it crosses the internet","Connect the virtual networks through an ExpressRoute circuit purchased from a connectivity provider","Peer the two virtual networks, so traffic stays on the Microsoft backbone network without public internet or gateways","Route the traffic through the on-premises datacenter by joining both virtual networks to the company's wide-area network"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/virtual-network/virtual-network-peering-overview","referenceTitle":"Azure Virtual Network Peering — Microsoft Learn"}}
{"id":"RAN-AZ900-00138","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7275954333d79786db28de9a683224e7c4c2f77eb62563b7fc05dbe65c4ff604","aliases":[],"snapshot":{"statement":"A company registers a new public domain name for a product launch. The team wants to host the domain's DNS zone on Azure's global network of name servers and manage its DNS records with the same credentials, APIs, tools, and billing they already use for their other Azure resources. Which Azure service should they use?","options":["Azure Virtual Network peering, which connects two virtual networks","Azure VPN Gateway, which sends encrypted traffic over the public internet","Azure DNS, which hosts DNS zones and resolves names","Azure ExpressRoute, which provides private connectivity to the Microsoft cloud"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-networking-services/5-domain-name-system/","referenceTitle":"Describe Azure DNS — Microsoft Learn"}}
{"id":"RAN-AZ900-00139","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:689d434dbbb51fd2d38ffa81d49e8d21ed5414f7e6b31931b36413feb9a76b86","aliases":[],"snapshot":{"statement":"A company wants its virtual machines to be reachable by custom names like payments.contoso.internal that resolve only inside its Azure virtual network and cannot be resolved from the public internet. Which Azure DNS capability supports this?","options":["Azure private DNS zones, which let the company use custom domain names in its private virtual networks instead of Azure-provided names","Public DNS zones with restricted access rules, which are meant to hide internet-facing records from queries outside the subscription","An ExpressRoute circuit, which is a private connection that keeps all name-resolution traffic off the public internet","A VPN gateway with DNS forwarding, which resolves private names for devices connected through its tunnels"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-networking-services/5-domain-name-system/","referenceTitle":"Describe Azure DNS — Microsoft Learn"}}
{"id":"RAN-AZ900-00140","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d2b1b80febd9d993df96212f3fd52efdd392d2cf54c2cfb0be6098376a7df3b7","aliases":[],"snapshot":{"statement":"A manufacturer must connect its on-premises datacenter network to the company's Azure virtual network. The link must be established quickly and cost-effectively, the team accepts that traffic crosses the public internet, and all data must travel inside an encrypted tunnel between the company's VPN device and Azure. Which Azure service and connection type should the company use?","options":["Azure ExpressRoute, with a private circuit ordered from a connectivity provider","Azure DNS, with a zone delegated to the on-premises DNS servers","Virtual network peering, with traffic kept on the Microsoft backbone network","Azure VPN Gateway, with a site-to-site connection to the on-premises VPN device"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/vpn-gateway/vpn-gateway-about-vpngateways","referenceTitle":"About Azure VPN Gateway — Microsoft Learn"}}
{"id":"RAN-AZ900-00141","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d0ed2fa774db3dda6e37a290da3bff762d89ca0fb3a78d6d9f661934b2b590c4","aliases":[],"snapshot":{"statement":"A consulting firm's analysts work from coworking spaces, client sites, and home, each using a company laptop. Each analyst needs secure access to resources inside the firm's Azure virtual network, as if the laptop were plugged into the office network, without any special device at the analysts' locations. Which VPN Gateway connection type fits?","options":["A site-to-site connection, which links the firm's office VPN device to the virtual network","A point-to-site connection, which connects individual devices to the virtual network over an encrypted tunnel","An ExpressRoute circuit, which connects each laptop through a dedicated private circuit","Virtual network peering, which connects the analysts' home networks to the virtual network"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/vpn-gateway/vpn-gateway-about-vpngateways","referenceTitle":"About Azure VPN Gateway — Microsoft Learn"}}
{"id":"RAN-AZ900-00142","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:15869ec35b59f95a8ba0bb49a25ad758ad839334678002581f2b997935db1520","aliases":[],"snapshot":{"statement":"A bank's trading platform in Azure must exchange market data with the bank's on-premises systems. The compliance team mandates that the connection never traverses the public internet, and the bank requires more reliability, consistent latencies, and higher security than typical internet connections. The bank is willing to work with a connectivity provider to establish the link. Which Azure service meets these requirements?","options":["Azure ExpressRoute, which extends the on-premises network into the Microsoft cloud over a private connection","Azure VPN Gateway, which encrypts traffic in a tunnel that crosses the public internet","Azure Virtual Network peering, which connects the bank's virtual network to the on-premises network over the Microsoft backbone","Azure DNS with private zones, which keeps the bank's traffic inside Azure during name resolution"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/expressroute/expressroute-introduction","referenceTitle":"Azure ExpressRoute Overview: Connect over a private connection — Microsoft Learn"}}
{"id":"RAN-AZ900-00143","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:803f7de0b2f37ad235212294c224135e2fe7a6154ce1d8e220b6b4c7e8e3c1fe","aliases":[],"snapshot":{"statement":"A retailer's application runs in an Azure virtual network and stores product images in an Azure storage account. The security team requires the application to reach the storage account through a private address inside the virtual network, keeping that traffic off the public internet, instead of through the storage account's public address. Which Azure feature accomplishes this?","options":["A public IP address attached to the storage account, so the application can reach it from anywhere on the internet","A network security group on the application's subnet, which filters traffic sent to the storage account's public address","A private endpoint, which gives the storage account a network interface with a private IP address from the virtual network","A VPN gateway between the application's subnet and the storage account, creating an encrypted tunnel over the internet"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/private-link/private-endpoint-overview","referenceTitle":"What is a private endpoint? — Microsoft Learn"}}
{"id":"RAN-AZ900-00144","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:30935fe429f483ef317d23d48731bc47e8dcdb56655d97f53186f6b7fe982cad","aliases":[],"snapshot":{"statement":"A SaaS provider is onboarding new engineers and needs to explain how Azure virtual networking classifies endpoints. Which statement describes a public endpoint?","options":["An endpoint that exists within a virtual network and has a private IP address from that network's address space","An endpoint that is reachable only from virtual networks peered with the provider's network","An endpoint that requires an ExpressRoute circuit before external clients can connect to it","An endpoint that has a public IP address and can be accessed from anywhere in the world"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-networking-services/2-virtual-network/","referenceTitle":"Describe Azure virtual networking — Microsoft Learn"}}
{"id":"RAN-AZ900-00145","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:328d36954a8da08c3632b63a7892437c55b29ec9391b577abd23066d4398d4db","aliases":[],"snapshot":{"statement":"A software company publishes a partner-facing REST API on an Azure service. External partners consume the API from their own systems over the internet, and none of them have access to the company's Azure virtual network. For the partners' access, which endpoint configuration is the appropriate fit?","options":["A private endpoint inside the company's virtual network, so partners connect through a private IP address","The service's public endpoint, which has a public IP address reachable from anywhere in the world","An Azure ExpressRoute circuit for each partner, keeping all partner traffic off the public internet","A point-to-site VPN for each partner, so partners tunnel into the company's virtual network"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-networking-services/2-virtual-network/","referenceTitle":"Describe Azure virtual networking — Microsoft Learn"}}
{"id":"RAN-AZ900-00146","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:a9ab1a0754df9a5ca587d8869f11862ec3052a120d5b0bd00313d5182227d2b8","aliases":[],"snapshot":{"statement":"A streaming service stores promotional artwork and episode video files that are delivered directly to viewers' browsers over HTTPS. The files are unstructured, and the platform must store very large volumes of them at low cost. Which Azure Storage service should the company use?","options":["Azure Blob Storage, an object store optimized for very large volumes of unstructured data","Azure Files, which provides fully managed file shares that clients access through the SMB and NFS protocols","Azure Disk Storage, which provides block-level storage volumes that are attached to virtual machines","Azure Queue Storage, which stores messages that application components exchange for asynchronous processing"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/storage/blobs/storage-blobs-introduction","referenceTitle":"Introduction to Blob (object) Storage — Microsoft Learn"}}
{"id":"RAN-AZ900-00147","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:10f029fdbd362b176e937770d4050ec1d81bc043316cd0aae95cce8f27408243","aliases":[],"snapshot":{"statement":"A law firm is moving its document-management application to Azure. The application expects a shared folder that several Windows virtual machines can mount simultaneously over SMB, and the firm does not want to build or maintain its own file server. Which Azure Storage service meets this requirement?","options":["Azure Blob Storage, with a container mounted as a shared folder on every virtual machine in the farm","Azure Disk Storage, attaching the same managed disk to each virtual machine so they share one drive","Azure Files, which provides fully managed cloud file shares that multiple cloud and on-premises clients can mount concurrently","Azure Table Storage, which stores each document as a schemaless key/attribute entity"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/storage/files/storage-files-introduction","referenceTitle":"What is Azure Files? — Microsoft Learn"}}
{"id":"RAN-AZ900-00148","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ede60ebf4c104f93da0e71e5710f361622c1fd2c1350587505dfe56aac1bc68f","aliases":[],"snapshot":{"statement":"An online retailer is decoupling its checkout flow. The web front end must record each incoming order so that a background worker can process it later; if the worker is briefly offline, no order may be lost. Which Azure Storage service is designed for this pattern?","options":["Azure Table Storage, writing each order as a NoSQL entity that the worker queries","Azure Queue Storage, which stores messages that the front end enqueues and the worker processes asynchronously","Azure Blob Storage, appending each order to a growing log blob that the worker follows","Azure Files, writing each order to a text file on a shared SMB folder"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/storage/queues/storage-queues-introduction","referenceTitle":"Introduction to Azure Queue Storage — Microsoft Learn"}}
{"id":"RAN-AZ900-00149","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:b56593a87c2c325c7dc9fd59cacd60eb56842f176b16d19fbb644f8664ad0efa","aliases":[],"snapshot":{"statement":"A startup is designing a consumer app that must be available worldwide. Beyond schemaless data, the team requires a fully managed NoSQL database with turnkey global distribution and single-digit-millisecond response times. A teammate proposes Azure Table storage, since it is already covered by their storage account. Which assessment is correct?","options":["Azure Table storage is the right choice, because it provides turnkey global distribution as part of a storage account","Azure Table storage can meet the latency target if the app uses one storage account per world region","Azure Blob Storage is the better option, because blobs are automatically replicated worldwide and read with low latency","Azure Cosmos DB is the better fit: a fully managed NoSQL database with turnkey global distribution and single-digit-millisecond response times"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/storage/tables/table-storage-overview","referenceTitle":"Introduction to Table storage - Object storage in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00150","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6a4738c2ed074941e9ae749815391f3b7e1cf202249878e239b502aa7b634b61","aliases":[],"snapshot":{"statement":"A manufacturing company runs its inventory database on an Azure virtual machine. The database needs persistent, block-level storage volumes for its operating system and data files, provisioned and managed by Azure. Which Azure Storage service should they use?","options":["Azure Blob Storage, storing the database files as objects in a container","Azure Files, mounting an SMB file share as the database volume","Azure Disk Storage, which provides block-level storage volumes for Azure virtual machines","Azure Queue Storage, which buffers database writes as messages between components"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/virtual-machines/managed-disks-overview","referenceTitle":"Overview of Azure Disk Storage — Microsoft Learn"}}
{"id":"RAN-AZ900-00151","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:04ca26a9eb208a0628fb794f0fb11e4aaa25ce36833cbb5ef29634f4c895b073","aliases":[],"snapshot":{"statement":"A web team already operates a standard general-purpose v2 storage account. For a new feature, they need to persist flexible, schemaless user-preference records and want to keep using their existing account rather than provisioning a new service. Which Azure Storage service should they use?","options":["Azure Blob Storage, saving each user's preferences as a JSON object in a container","Azure Table Storage, a schemaless key/attribute store that lives inside their existing storage account","Azure Queue Storage, storing each preference record as a message until it is read","Azure Disk Storage, dedicating a premium storage volume to the preferences data"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/storage/tables/table-storage-overview","referenceTitle":"Introduction to Table storage - Object storage in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00152","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:199f1dc0cf4142d69a6b25efb1593ee3a52066841c6946d39879e72998fc1495","aliases":[],"snapshot":{"statement":"A news website stores the images displayed in its articles. Readers load these images continuously throughout the day, and every page view triggers reads. Which access tier keeps the total cost of storing and reading this data lowest?","options":["The hot tier, which has the highest storage cost but the lowest access cost","The cool tier, which has lower storage costs and lower access costs than hot","The cold tier, which is optimized for data that is rarely accessed","The archive tier, which has the lowest storage cost of any tier"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/storage/blobs/access-tiers-overview","referenceTitle":"Access tiers for blob data — Microsoft Learn"}}
{"id":"RAN-AZ900-00153","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:10674c3d32dc18215ecdf0a027687a3b71b534ea14f1dd8a473d71463bfad2a8","aliases":[],"snapshot":{"statement":"A utility company must retain meter-reading exports for 90 days. During that period the exports are read only rarely, mainly to investigate billing disputes, but whenever a dispute arrives the data must be readable immediately. The team wants the lowest storage cost that still allows instant reads. Which access tier is the best fit?","options":["The archive tier, because it has the lowest storage cost of any tier and disputes are rare","The cool tier, because it is an online tier for infrequently accessed data with a 30-day minimum storage duration","The hot tier, because instantly available data must always be stored in the highest tier","The cold tier, an online tier for rarely accessed data with fast retrieval and a 90-day minimum storage duration"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/storage/blobs/access-tiers-overview","referenceTitle":"Access tiers for blob data — Microsoft Learn"}}
{"id":"RAN-AZ900-00154","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:426f35b8de58d42d3057d426befcfada4f46d8c6b185a7d1f0efd338ba573506","aliases":[],"snapshot":{"statement":"A hospital must retain seven years of audit records for compliance purposes. The records are almost never opened; when an audit occurs, the compliance team can tolerate waiting several hours before the files are available. The hospital wants to minimize storage cost. Which access tier should they choose?","options":["The cool tier, an online tier for infrequently accessed data with a 30-day minimum storage duration","The archive tier, an offline tier with the lowest storage cost; blobs must be rehydrated before they can be read","The cold tier, an online tier that is cheaper to store in than archive and still serves instant reads","The hot tier, which minimizes access costs for the occasional audit reads"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/storage/blobs/access-tiers-overview","referenceTitle":"Access tiers for blob data — Microsoft Learn"}}
{"id":"RAN-AZ900-00155","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b81b4581fae097f65614bd982790b3beb78fdd5e299d72515a74eb469febf5f5","aliases":[],"snapshot":{"statement":"An analyst sees that the cool tier has a lower storage price than hot and proposes moving the company's product catalog images — which website visitors view constantly — to cool to cut spending. Why would this change most likely increase the total cost?","options":["Cool has lower storage costs but higher access costs than hot, and constant reads can outweigh the savings","Cool blobs can only be read through authenticated REST calls, which are billed at a premium","Cool storage is offline, so every read would first require a rehydration that takes hours","Cool requires data to be stored for at least 180 days, and the product catalog changes every week"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/storage/blobs/access-tiers-overview","referenceTitle":"Access tiers for blob data — Microsoft Learn"}}
{"id":"RAN-AZ900-00156","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:9fb1e4872e166984c789590d59a46f31bef74f677b4910bdebaad9e3a921c4c4","aliases":[],"snapshot":{"statement":"An insurance company stores policy documents in blob storage. Customers download their documents heavily during the first month after issue; after that, documents are retrieved only a few times a year, but every request must be served immediately when a customer logs in. Documents are kept for many years. Which arrangement meets the requirements at the lowest cost?","options":["Move every document to the archive tier at upload, because archive is an online tier with the lowest storage cost","Move documents older than one month to the archive tier, rehydrating them whenever a customer logs in","Keep recent documents in the hot tier and move older documents to cool or cold, which remain online and instantly available","Keep all documents in the hot tier forever, because a blob's access tier can no longer be changed after upload"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/4-describe-azure-storage-services/","referenceTitle":"Describe Azure storage services — Microsoft Learn"}}
{"id":"RAN-AZ900-00157","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:d5a9e019e9e6cc946d508f00003e51e9fbba8481843c3cbf578b2764a1c52498","aliases":[],"snapshot":{"statement":"A small design studio is creating its first storage account and wants the lowest-cost redundancy option. The founders accept that a disaster affecting an entire data center could cause data loss, as long as the data is still replicated three times against routine hardware failures. Which option should they choose?","options":["Locally redundant storage (LRS), which replicates data three times within a single data center","Zone-redundant storage (ZRS), which replicates data synchronously across three availability zones in the primary region","Geo-redundant storage (GRS), which replicates data three times and also copies it asynchronously to a secondary region","Geo-zone-redundant storage (GZRS), which combines availability zone replication with a secondary region"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/3-redundancy/","referenceTitle":"Describe Azure storage redundancy — Microsoft Learn"}}
{"id":"RAN-AZ900-00158","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d74f89ba25663a04c67d3cd6c060783673259de32479bd39f68e345ba870efbc","aliases":[],"snapshot":{"statement":"A financial services firm runs its storage in an Azure region that supports availability zones. Regulators require that data remains available for both reads and writes even if one availability zone fails, but all copies must stay within the primary region. Which redundancy option should the firm use?","options":["Locally redundant storage (LRS), which keeps three synchronized copies in a single data center","Zone-redundant storage (ZRS), which replicates data synchronously across three availability zones in the primary region","Geo-zone-redundant storage (GZRS), which replicates across zones and also to a secondary region","Read-access geo-redundant storage (RA-GRS), which serves reads from a copy in a secondary region"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/3-redundancy/","referenceTitle":"Describe Azure storage redundancy — Microsoft Learn"}}
{"id":"RAN-AZ900-00159","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:98ee203b26485be4684ffd8348cb12714b0f9937aa4f06d177e92689b0a15296","aliases":[],"snapshot":{"statement":"A logistics company stores shipment records in a storage account configured with geo-redundant storage (GRS). During an audit, the compliance officer asks exactly where the data copies live. Which description of GRS is accurate?","options":["Three copies spread across three availability zones of the primary region, with no copies elsewhere","Three synchronous copies in the primary region, plus a single readable copy in the secondary region at any time","Three synchronous copies in the primary region, then asynchronous replication to a secondary region, where three copies are stored again","Nine copies evenly distributed across the availability zones of both regions"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/3-redundancy/","referenceTitle":"Describe Azure storage redundancy — Microsoft Learn"}}
{"id":"RAN-AZ900-00160","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:fac9a76bf9de166c8c6cc39887005b2fbd1fed82e69ef36eebb259c79701a5f3","aliases":[],"snapshot":{"statement":"A healthcare platform runs in an availability-zone-enabled region. The platform must tolerate the failure of an individual availability zone and also survive a full regional outage, keeping a durable copy in a second region. Which redundancy option was designed for exactly this combination?","options":["Zone-redundant storage (ZRS), which survives zone failures but keeps all copies in the primary region","Geo-redundant storage (GRS), which survives regional outages but keeps the primary copies in a single data center","Read-access geo-redundant storage (RA-GRS), which combines zone replication with a readable secondary region","Geo-zone-redundant storage (GZRS), which replicates across three availability zones in the primary region and to a secondary region"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/3-redundancy/","referenceTitle":"Describe Azure storage redundancy — Microsoft Learn"}}
{"id":"RAN-AZ900-00161","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:c00b7219523ee9d4ade9eb8bae1bd8bddf6d5f08a6385a7488bea8aebe658afb","aliases":[],"snapshot":{"statement":"A travel company's storage account uses geo-redundant storage (GRS). The architecture team wants to serve part of the application's read traffic from the copy in the secondary region during normal operations, without waiting for a disaster. What should the team do?","options":["Nothing — under GRS, applications can already read the secondary region's copy at any time","Use read-access geo-redundant storage (RA-GRS), which makes the replicated data in the secondary region available for reads","Trigger a storage account failover to the secondary region, then fail back after each batch of reads","Switch to zone-redundant storage (ZRS), which serves reads from a second availability zone in the same region"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/3-redundancy/","referenceTitle":"Describe Azure storage redundancy — Microsoft Learn"}}
{"id":"RAN-AZ900-00162","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6d9465d4b6a0d58b0b76550a2be1c59144fefe470cceb34519b88b5499c68078","aliases":[],"snapshot":{"statement":"During onboarding, a new cloud engineer asks the storage team to explain, in one sentence, what an Azure storage account fundamentally provides within the platform. Which explanation is correct?","options":["A unique namespace for Azure Storage data; the account name is part of every object's address","A virtual machine in Azure that hosts your blobs and files on managed disks","A network boundary that isolates storage services from the rest of the subscription","A replication policy that determines how many copies of your data are kept and in which regions"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/2-accounts/","referenceTitle":"Describe Azure storage accounts — Microsoft Learn"}}
{"id":"RAN-AZ900-00163","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:8535671594c57c525c4cdc6c9dbeb8affee0f7becf01f99b5ad4b56431a97068","aliases":[],"snapshot":{"statement":"A software company needs a single storage account for a mixed workload: blobs for user uploads, file shares for a legacy tool, queues for inter-component messages, and tables for lightweight records. Which storage account type should they create?","options":["A premium block blobs account, which serves all four storage services on solid-state drive hardware","A premium file shares account, which hosts blobs, files, queues, and tables for high-performance apps","A standard general-purpose v2 account, supporting Blob, Queue, Table, and Azure Files — recommended for most scenarios","A premium page blobs account, which stores every storage service as page blobs"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/storage/common/storage-account-overview","referenceTitle":"Storage Account Overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00164","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3888d7bb0d4b50ad4e4dc20e6343b7cc7dfd9c2b28cbb1b682b161ed40cc1ced","aliases":[],"snapshot":{"statement":"An IoT platform writes millions of small sensor readings every hour into block blobs. The ingestion pipeline requires consistently low storage latency, and the team is willing to pay for solid-state drive performance. Which storage account type is recommended for this workload?","options":["A standard general-purpose v2 account, which offers the lowest per-gigabyte capacity prices for Azure Storage","A premium file shares account, which provides SSD-backed storage for high-performance file shares","A premium page blobs account, which is designed for workloads that use page blobs","A premium block blobs account, recommended for high transaction rates and consistently low storage latency"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/storage/common/storage-account-overview","referenceTitle":"Storage Account Overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00165","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:776cfc8d534511ee13ade4f625f5ddef54f6cdb9b22ff826b5ca4808cef96f8e","aliases":[],"snapshot":{"statement":"A media company's rendering application reads source footage from an Azure file share and needs SSD-backed performance to keep render jobs fast. The team also wants the same storage account to serve both SMB and NFS file shares. Which storage account type should they use?","options":["A premium file shares account, which supports SSD-backed file shares over both SMB and NFS","A standard general-purpose v2 account, which supports NFS file shares alongside its standard performance tier","A premium block blobs account, which serves files over SMB and NFS from blob storage","A premium page blobs account, which converts page blobs into SMB file shares"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/storage/common/storage-account-overview","referenceTitle":"Storage Account Overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00166","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:a22e42e187cfd8b2d52dcdf91bc465c41516d3af424c7966e2df8bb19425cd9f","aliases":[],"snapshot":{"statement":"A furniture retailer has run a legacy Blob storage account for years. A new project now needs to add an Azure file share and a queue in the same account, reusing the existing namespace. What is the recommended path?","options":["Enable the Files and Queues features in the legacy account's configuration settings, which are currently disabled","Upgrade the legacy account to general-purpose v2, which supports Blob, Queue, Table, and Azure Files","Create a premium block blobs account and move the file shares and queues into it for SSD performance","Create a premium file shares account, which supports blobs, queues, and tables at SSD performance"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/storage/common/storage-account-upgrade","referenceTitle":"Upgrade to a general-purpose v2 storage account — Microsoft Learn"}}
{"id":"RAN-AZ900-00167","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:3d7edb8499a7f774d7d1f3d218d4cd692f70a6d6ff8eca10177e48c0ec3547d9","aliases":[],"snapshot":{"statement":"An operations team wants to automate a nightly upload of backup files from an on-premises Linux server to Azure Blob Storage. The job must run unattended as the final step of an existing shell script. Which tool should they use?","options":["Azure Storage Explorer, a standalone desktop app with a graphical interface for managing storage data","Azure File Sync, which synchronizes a Windows Server with an Azure file share","AzCopy, a command-line utility for copying blobs or files to and from a storage account","Azure Migrate, which assesses and migrates on-premises servers to Azure"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/storage/common/storage-use-azcopy-v10","referenceTitle":"Copy or move data to Azure Storage by using AzCopy v10 — Microsoft Learn"}}
{"id":"RAN-AZ900-00168","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:504621bcc213c0f3bf41e67c88a844e0b6ffbcd4ddc2763fcc1394ff3515e0c6","aliases":[],"snapshot":{"statement":"A marketing analyst regularly uploads campaign artwork, downloads performance reports, and reorganizes folders across several of the company's storage accounts. She is not comfortable with command-line tools and wants a visual interface on her laptop. Which tool should she use?","options":["Azure Storage Explorer, a standalone app with a graphical interface for managing storage data","AzCopy, which copies blobs and files between machines and storage accounts from the command line","Azure File Sync, which caches the storage accounts on her Windows laptop","Azure Data Box, which ships physical devices with storage data to and from her office"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/7-identify-azure-file-movement-options/","referenceTitle":"Identify Azure file movement options — Microsoft Learn"}}
{"id":"RAN-AZ900-00169","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d49ff5cd0fca2646265c62748c689964708bb0b846e7a5ca96e8d817f7ab0864","aliases":[],"snapshot":{"statement":"A manufacturer has branch offices on several continents. The IT team wants to centralize the company's file shares in Azure while keeping fast local access at each branch, caching the most frequently used files on the existing Windows Servers. Which approach should they use?","options":["Scheduled AzCopy jobs that copy the entire share from Azure to every branch server every night","Azure Storage Explorer installed on each branch server, which keeps the local folders synchronized with Azure","Azure Data Box shipments that deliver a fully updated copy of the share to each branch office on request","Azure File Sync, which syncs Windows Servers with an Azure file share and caches frequently accessed files locally"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/storage/file-sync/file-sync-introduction","referenceTitle":"What is Azure File Sync? — Microsoft Learn"}}
{"id":"RAN-AZ900-00170","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:8679331e0b662790be397d549dc8b553fc2d80eba3a23c02a85a37132b7ca001","aliases":[],"snapshot":{"statement":"A logistics company acquires a smaller firm that keeps its product catalog — thousands of images and PDFs — in Amazon S3. As part of the consolidation, the cloud team must copy that content into the company's Azure Blob Storage account. Which tool should the team use to move the files between the two cloud providers?","options":["Azure Storage Explorer, which manages files and blobs in Azure Storage accounts through its graphical interface","AzCopy, which can be configured to work with other cloud providers to copy data into Azure Storage","Azure File Sync, which keeps a Windows Server bi-directionally synchronized with an Azure file share","Azure Data Box, which ships a physical device so bulk data can be loaded locally and returned for upload"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/7-identify-azure-file-movement-options/","referenceTitle":"Identify Azure file movement options — Microsoft Learn"}}
{"id":"RAN-AZ900-00171","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:da854c79624871216acd310cfe12ab5391fb056ef75b930f8f7a0945beda8575","aliases":[],"snapshot":{"statement":"During a design review for a hybrid file solution, the team makes four statements about the tools that move files and data in Azure. Which statement is accurate?","options":["When AzCopy synchronizes blobs or files, it copies in one direction only, from the designated source to the destination","Azure Storage Explorer keeps a local Windows Server continuously and automatically synchronized with an Azure file share","Azure File Sync is required for any upload performed through Azure Storage Explorer","Azure Data Box uploads files to Azure over the customer's internet connection while the device is on site"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/7-identify-azure-file-movement-options/","referenceTitle":"Identify Azure file movement options — Microsoft Learn"}}
{"id":"RAN-AZ900-00172","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:08f320097f888f68ab1ee2edcd10f5e4c565278a7bf5cf940afabe3b4862433d","aliases":[],"snapshot":{"statement":"A public sector agency plans to move its on-premises datacenter — servers, databases, and web applications — to Azure. Leadership wants a central place to assess readiness and hosting costs before deciding, and then run the migration with minimal downtime. Which service should they use?","options":["Azure Data Box, which ships a physical appliance to the customer to transfer bulk data into Azure","Azure File Sync, which extends on-premises file servers into Azure file shares","Azure Migrate, a central hub for assessing and migrating on-premises servers, databases, and apps to Azure","AzCopy, which copies server disks into storage accounts over the network"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/migrate/migrate-services-overview","referenceTitle":"About Azure Migrate — Microsoft Learn"}}
{"id":"RAN-AZ900-00173","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ac9c81280ce07df1018e7a7032902fdfc4273b39e01994d902a0c8b8b5795969","aliases":[],"snapshot":{"statement":"A film studio must move tens of terabytes of raw footage into Azure before an archive deadline. The studio's internet connection is too slow to transfer the data over the network within the available window. Which option is designed for this situation?","options":["AzCopy, running continuously over the studio's existing internet connection until the whole transfer completes","Azure Migrate, which transfers large media archives to Azure over an accelerated network link","Azure File Sync, which uploads the footage to Azure in the background as it is edited","Azure Data Box, a physical migration service where a shipped device is loaded locally and returned for upload"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/databox/data-box-overview","referenceTitle":"Microsoft Azure Data Box overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00174","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:da2c1c10ed2bc9eaf235b159321e7a9bbfd149cfeb525f80f7a2b36b0d6a6312","aliases":[],"snapshot":{"statement":"An IT administrator at a research institute has just taken delivery of an Azure Data Box device at the loading dock. Which sequence describes how the import process works, from setup to upload?","options":["Microsoft permanently installs the device in the datacenter as an on-premises edge cache for Azure Storage","The administrator copies data onto the device, which uploads it to Azure over the internet while it is on site","The administrator ships the institute's own storage disks to Microsoft, which loads the data into a storage account","The device is set up through its local web UI, loaded on site, and returned; Microsoft uploads the data to Azure"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/6-identify-azure-data-migration-options/","referenceTitle":"Identify Azure data migration options — Microsoft Learn"}}
{"id":"RAN-AZ900-00175","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:81117db971f3d30ee8870e80c1dd5d79388bf483ff27aaf8ba3faec17f8a40e7","aliases":[],"snapshot":{"statement":"During a datacenter exit, a grocery chain must accomplish two things: evaluate which on-premises servers are ready for Azure and migrate them with minimal downtime, and move an archive of tens of terabytes of historical sales data over a slow WAN link within a fixed window. Which combination of services should they use?","options":["Azure Data Box for both, because it also assesses server readiness before the data is shipped","Azure Migrate for both, because its online data transfers are unaffected by the slow WAN link","Azure Migrate to assess and migrate the servers, and Azure Data Box to move the archive","AzCopy and Azure File Sync, which together provide server assessment and physical data transfer"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-storage-services/6-identify-azure-data-migration-options/","referenceTitle":"Identify Azure data migration options — Microsoft Learn"}}
{"id":"RAN-AZ900-00176","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:85c19a4cffc0e4e7456df3688b888fdb26864bb02d2148e65b3e67eb8839f440","aliases":[],"snapshot":{"statement":"A regional hospital network is moving its email, collaboration tools, and several custom web apps to the cloud. The IT director wants a single cloud service that manages employee identities and lets each employee sign in with one account to reach Microsoft 365 and the custom apps.\nWhich Azure service should the director choose?","options":["Microsoft Entra Domain Services, which runs managed domain controllers for applications that need classic AD DS","Azure Firewall, which filters network traffic between the hospital's apps and the internet","Microsoft Entra ID, the cloud-based identity and access management service for Microsoft and custom cloud applications","Microsoft Purview, which governs and classifies the hospital's data estate"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/2-directory-services","referenceTitle":"Describe Azure directory services — Microsoft Learn"}}
{"id":"RAN-AZ900-00177","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:027737e9368b15625cd517ea211b0cf997a9ffe110d12b89448a2e775e31718b","aliases":[],"snapshot":{"statement":"An industrial-equipment manufacturer runs a legacy inventory application that authenticates users with NTLM and needs its servers to be domain-joined. The company is moving the application to Azure and wants to retire its on-premises domain controllers entirely, with no configuring, patching, or monitoring of domain controllers by its own staff.\nWhich approach meets these requirements?","options":["Deploy two Windows Server virtual machines as domain controllers and run them yourself in Azure","Create a Microsoft Entra Domain Services managed domain, in which Microsoft runs the domain controllers","Use Microsoft Entra ID alone, since it provides NTLM authentication and domain join for the application","Keep the on-premises domain controllers and extend them to Azure with a site-to-site VPN"],"answer":"B","reference":"https://learn.microsoft.com/en-us/entra/identity/domain-services/overview","referenceTitle":"Overview of Microsoft Entra Domain Services — Microsoft Learn"}}
{"id":"RAN-AZ900-00178","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f981af088946191b17fa405c460e93c47e190819f9eded078de429c44e586fc4","aliases":[],"snapshot":{"statement":"A grocery chain uses a Microsoft Entra Domain Services managed domain so that a legacy warehouse application can keep using Kerberos and domain join. A newly hired identity administrator asks how employee identities relate to the managed domain: where the accounts live and how they reach the managed domain.\nWhich response is accurate?","options":["The managed domain synchronizes two ways with Microsoft Entra ID, so changes made in either directory appear in both","The managed domain keeps its own user accounts, and employees receive separate credentials for domain-joined machines","The managed domain pulls accounts directly from an on-premises Active Directory forest, bypassing Microsoft Entra ID","The managed domain is fed by a one-way synchronization from Microsoft Entra ID, the source of authority for identities"],"answer":"D","reference":"https://learn.microsoft.com/en-us/entra/identity/domain-services/overview","referenceTitle":"Overview of Microsoft Entra Domain Services — Microsoft Learn"}}
{"id":"RAN-AZ900-00179","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:f1b443dbc6b15b468742d53e66f70fd9c857a99f995ac8b1a8249522c1c78f08","aliases":[],"snapshot":{"statement":"During onboarding at a public-sector agency, a junior cloud architect is asked to summarize, in one sentence, how Microsoft Entra ID and Microsoft Entra Domain Services divide the work in Azure.\nWhich statement is accurate?","options":["Microsoft Entra Domain Services provides managed domain services (domain join, group policy, LDAP, Kerberos/NTLM authentication) without any domain controllers to manage","Microsoft Entra ID provides domain join, group policy, LDAP, and Kerberos/NTLM authentication for legacy line-of-business applications","Microsoft Entra Domain Services is the identity service that employees use to sign in to Microsoft 365, SaaS applications, and the company's own custom cloud applications","Microsoft Entra Domain Services requires you to deploy, patch, and monitor your own Windows Server domain controllers in Azure"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/2-directory-services","referenceTitle":"Describe Azure directory services — Microsoft Learn"}}
{"id":"RAN-AZ900-00180","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:b9b5fb30986eea2116a027e17692e0eb467fba143ea9386b4ef398f6837b7907","aliases":[],"snapshot":{"statement":"At a logistics company, warehouse staff must sign in separately to the route-planning app, the inventory app, and Microsoft 365, and the help desk handles a constant stream of password resets. The IT lead wants employees to sign in once with one set of credentials and then open every assigned app without signing in again.\nWhich capability should the IT lead enable?","options":["Multifactor authentication, prompting each user for an additional verification factor during the sign-in process","Passwordless sign-in, replacing passwords with Windows Hello or a FIDO2 security key","Single sign-on, which lets users sign in once and reach every assigned application","Conditional Access policies, blocking or challenging sign-ins that come from unexpected locations"],"answer":"C","reference":"https://learn.microsoft.com/en-us/entra/identity/enterprise-apps/what-is-single-sign-on","referenceTitle":"What is single sign-on in Microsoft Entra ID? — Microsoft Learn"}}
{"id":"RAN-AZ900-00181","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:852f6913f9f3b7e76521b7fc0e2c0a38addfe5ea66273664d058c2180e12e59c","aliases":[],"snapshot":{"statement":"A payments company suffered a phishing incident in which attackers captured employees' passwords. The security team now requires each sign-in to demand more than a password: after entering the password, each employee must approve the request on their registered phone before access is granted.\nWhich approach does the team require at sign-in?","options":["Multifactor authentication, combining something the employee knows with something the employee has","Single sign-on, using one trusted authentication to cover every application","Passwordless authentication, replacing the password with a biometric or a hardware security key","Longer passwords rotated more frequently, so that a stolen password becomes obsolete quickly"],"answer":"A","reference":"https://learn.microsoft.com/en-us/entra/identity/authentication/concept-mfa-howitworks","referenceTitle":"Microsoft Entra multifactor authentication overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00182","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:535be7f01ba3b2e4dad59a5eed5a442e8eae4b7332094a2a055408e27f49167b","aliases":[],"snapshot":{"statement":"A hotel chain's IT team spends too much effort on password resets, and auditors flag password reuse across properties. The team decides to remove passwords from the sign-in experience altogether: staff will sign in with a biometric or a PIN tied to their own Windows PC, with a tap of a FIDO2 security key, or with a number match in the Microsoft Authenticator app.\nWhich approach is the team adopting?","options":["Multifactor authentication, which adds an extra verification factor on top of the user's password at sign-in","Single sign-on, which lets one authentication cover all of the chain's applications","Conditional Access, which requires a second factor only for sign-ins from unexpected locations","Passwordless authentication, which eliminates passwords with Windows Hello for Business, Authenticator, or FIDO2 security keys"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/3-authentication-methods","referenceTitle":"Describe Azure authentication methods — Microsoft Learn"}}
{"id":"RAN-AZ900-00183","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:fc5b31face28a8d9381e4ea754564f7af3cc0d9b79843d2497406c74aa6408a5","aliases":[],"snapshot":{"statement":"A junior security analyst at a media company is writing study notes about the authentication methods available in Microsoft Entra ID. The notes must state precisely what single sign-on, multifactor authentication, and passwordless sign-in each do.\nWhich statement should the notes contain?","options":["Single sign-on requires a hardware key at every sign-in; multifactor authentication covers one application at a time; passwordless still asks for the password on untrusted devices","Single sign-on lets a user sign in once and access multiple applications; multifactor authentication requires credentials from more than one category; passwordless methods eliminate the password","Single sign-on eliminates the password; multifactor authentication signs users in with no credential at all; passwordless adds a second factor to the password","Single sign-on applies to Microsoft apps only; multifactor authentication replaces passwords with biometrics; passwordless rotates passwords automatically"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/3-authentication-methods","referenceTitle":"Describe Azure authentication methods — Microsoft Learn"}}
{"id":"RAN-AZ900-00184","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:5d18fc954ae5d3476595ab8e8b95b6608684ce15dfff90c3bf1453d6e596d5dc","aliases":[],"snapshot":{"statement":"A utility company requires multifactor authentication for remote access to its control-room dashboard. A dispatcher signs in with her password and then approves the request on her phone, which is registered as her verification method.\nWhich two categories of credentials has the dispatcher presented?","options":["Something you know and something you are","Something you have and something you are","Something you know and something you have","Something you are, because the phone is unlocked with a fingerprint"],"answer":"C","reference":"https://learn.microsoft.com/en-us/entra/identity/authentication/concept-mfa-howitworks","referenceTitle":"Microsoft Entra multifactor authentication overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00185","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e3bbd8c1f76b824c54de3c5c1a68a340badd9eb1d560db7430d73bb7663d39bf","aliases":[],"snapshot":{"statement":"A university runs a joint research program with an engineering firm. About 40 of the firm's engineers need access to a few of the university's web apps for the semester. The firm manages its own identities, and its engineers must keep using their existing work accounts rather than new credentials. The university must stay in control of which apps the guests can reach.\nWhich approach should the university use?","options":["Create a local account in the university's tenant for each engineer, with new passwords issued by the university","Invite the engineers as guest users through Microsoft Entra B2B collaboration, so they sign in with their own work identities","Set up B2B direct connect with the firm's tenant, so the engineers can reach the university's web apps from their home tenant","Use External ID for customers, so the engineers sign up for the university's apps with social identities"],"answer":"B","reference":"https://learn.microsoft.com/en-us/entra/external-id/what-is-b2b","referenceTitle":"What is Microsoft Entra B2B collaboration? — Microsoft Learn"}}
{"id":"RAN-AZ900-00186","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d571d94bc73931b8fcfefdeb9c7415de2d75979e636cccbbf4a0cd261e895a8f","aliases":[],"snapshot":{"statement":"An environmental nonprofit invites staff from a partner agency to collaborate on its Microsoft 365 environment. The invited users have just redeemed their invitations.\nWhat does the nonprofit's administrator now see in the nonprofit's own directory?","options":["Guest user objects representing the partners, who sign in with their own work, school, or social identities","Nothing; invited users remain invisible until they register new credentials in the nonprofit's tenant","Full employee accounts with temporary passwords that the nonprofit must reset at first sign-in","A read-only replica of the partner agency's entire home directory, synchronized into the nonprofit's tenant"],"answer":"A","reference":"https://learn.microsoft.com/en-us/entra/external-id/what-is-b2b","referenceTitle":"What is Microsoft Entra B2B collaboration? — Microsoft Learn"}}
{"id":"RAN-AZ900-00187","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:25cbbc77354abb1a07f4722fb99023ab8a8972f76dbd753961f8a637d6f50446","aliases":[],"snapshot":{"statement":"An automotive supplier onboarding a new parts partner asks its security lead to explain, for the kickoff deck, how external identities work in Azure. Both companies already use Microsoft 365 for day-to-day collaboration.\nWhich statement is accurate?","options":["External users must be issued accounts with passwords in the supplier's tenant before they can access any resource","B2B collaboration requires partners to run Microsoft Entra ID on their side","External identities let outsiders access approved resources with their existing credentials, while the organization controls what they reach","External identities give partner users the same access to resources as the supplier's own employees"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/4-external-identities","referenceTitle":"Describe Azure external identities — Microsoft Learn"}}
{"id":"RAN-AZ900-00188","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:9c1ac9c6f0b1940e8d523365c2227b285aa1103809f1e26934a7cb4afaf127f7","aliases":[],"snapshot":{"statement":"A streaming service wants a consistent rule for its Microsoft 365 apps: employees signing in from the office or other known locations proceed normally, but a sign-in from an unexpected location must be challenged for a second factor before access is granted.\nWhich Microsoft Entra capability should the security team configure?","options":["Azure role-based access control, assigning the Contributor role at the subscription scope","Multifactor authentication enforced for every user on every sign-in, regardless of location","Azure Policy, blocking the deployment of any resources that do not match its compliance rules","Microsoft Entra Conditional Access, which evaluates sign-in signals and enforces decisions such as requiring MFA"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/5-conditional-access","referenceTitle":"Describe Azure conditional access — Microsoft Learn"}}
{"id":"RAN-AZ900-00189","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:db5dc4c76f7aa0e50e3f588f1a4cdea630e3e4de230b6dc225853ce8a85a6d50","aliases":[],"snapshot":{"statement":"An identity architect at a software company is sketching how Microsoft Entra Conditional Access works for a design review. The sketch has two boxes: one for the signals a policy can evaluate, one for the decisions it can enforce.\nWhich list is accurate?","options":["Signals — allow access, require multifactor authentication, block access; decisions — user or group, IP location, device in use, application, sign-in risk","Signals — virtual machine size, storage tier, and region; decisions — allow access, possibly requiring multifactor authentication, or block access","Signals — user or group, IP location, device in use, application, sign-in risk; decisions — create a resource group, assign the Reader role, delete a subscription","Signals — user or group, IP location, device in use, application, sign-in risk; decisions — allow access, possibly requiring multifactor authentication, or block access"],"answer":"D","reference":"https://learn.microsoft.com/en-us/entra/identity/conditional-access/overview","referenceTitle":"Microsoft Entra Conditional Access: Zero Trust Policy Engine — Microsoft Learn"}}
{"id":"RAN-AZ900-00190","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:59f3db049021c96c3f4b437800f7d262b937bbd6f2ec99a7921b812e53fb0229","aliases":[],"snapshot":{"statement":"During a design review at an e-commerce company, a newly hired architect keeps mixing up Conditional Access with Azure RBAC and Azure Policy.\nWhich statement would correctly describe Conditional Access?","options":["Conditional Access policies grant users permissions to create and manage Azure resources, such as virtual machines, storage accounts, and virtual networks","Azure RBAC role assignments are evaluated at sign-in and challenge users for a second factor before granting resource access","Conditional Access policies evaluate identity signals, such as user, location, and device, and can allow access, require MFA, or block","Conditional Access policies enforce tagging and allowed regions on resources, updating the rules after those resources are deployed"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/5-conditional-access","referenceTitle":"Describe Azure conditional access — Microsoft Learn"}}
{"id":"RAN-AZ900-00191","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:f6d1e27b0e1932b5a1e7cc783e853db3c45524ac00eefd2c718cb93c26943968","aliases":[],"snapshot":{"statement":"A SaaS company with several engineering teams must control access to its Azure resources. Each team should receive only the level of access it needs to complete its tasks — no more — and that access should attach to a specific set of resources, such as one resource group.\nWhich mechanism should the company use?","options":["Azure role-based access control, granting access through role assignments at a chosen scope","Microsoft Entra Conditional Access, evaluating identity signals at sign-in and challenging risky sign-ins","Azure Policy definitions, auditing and enforcing configuration rules on deployed resources","Resource locks, preventing the accidental deletion or modification of critical resources"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/6-role-based-access-control","referenceTitle":"Describe Azure role-based access control — Microsoft Learn"}}
{"id":"RAN-AZ900-00192","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3506ae73df61ebd472020894633d8a6cbab4c55f3032c0ee9a45f1ec05d9c267","aliases":[],"snapshot":{"statement":"A game studio's deployment team must be able to create and manage all of the Azure resources in the studio's \"game-builds\" resource group. The studio's security policy adds one restriction: the team must not be able to grant other people access to those resources.\nWhich built-in role should the team be assigned at the resource group scope?","options":["The Reader role, which grants access to view resources without allowing any changes","The Owner role, which grants full access to everything in the assigned scope","The User Access Administrator role, which lets the team manage user access to Azure resources","The Contributor role, which grants full access to manage all resources"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/role-based-access-control/built-in-roles","referenceTitle":"Azure built-in roles — Microsoft Learn"}}
{"id":"RAN-AZ900-00193","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:272b2196286ac62738240d5ab81a3eb92b373cfcb0c1e912d9a05b6ccfcef234","aliases":[],"snapshot":{"statement":"External auditors at a pharmaceutical company need to view every resource in one subscription — including resources in resource groups that will be created months from now — but they must not be able to change anything. The identity administrator wants a single role assignment that covers all of this automatically.\nWhat should the administrator do?","options":["Assign the Owner role to the auditors' group at the subscription scope, so that inheritance covers new resource groups","Assign the Reader role to the auditors' group at the subscription scope, so that inheritance covers every resource group and resource in the subscription","Assign the Reader role to the auditors' group separately in each existing resource group, repeating the assignment whenever a new resource group is created","Assign the Contributor role to the auditors' group at the management group scope, so that the whole hierarchy is covered"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/6-role-based-access-control","referenceTitle":"Describe Azure role-based access control — Microsoft Learn"}}
{"id":"RAN-AZ900-00194","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:4b3ee0ad9f0267811366d04376c61179b3ae0145086c673aedd383c2f9052d8e","aliases":[],"snapshot":{"statement":"A platform engineer is mentoring a newly hired Azure administrator on how role-based access control works. The administrator asks what a role assignment is made of and how permissions travel across the hierarchy.\nWhich statement should the engineer make?","options":["Role definitions must be created individually for each user, so every user ends up with a custom role","Access granted at a child scope is inherited upward by parent scopes, so a role assigned in a resource group also applies at the subscription level","Every role assignment combines a security principal, a role definition, and a scope, and access granted at a parent scope is inherited by child scopes","Role assignments are enforced at application sign-in, where they can challenge the user for a second factor"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/role-based-access-control/overview","referenceTitle":"What is Azure role-based access control (Azure RBAC)? — Microsoft Learn"}}
{"id":"RAN-AZ900-00195","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:3c148e2f3832a18da5a3d50d6d664fff963bec3adb45f727d5694eae14788e9a","aliases":[],"snapshot":{"statement":"An insurance carrier's security office publishes a new access policy: always authenticate and authorize every request based on all available data, grant only just-enough access for the task at hand, and plan as if attackers are already inside the environment.\nWhich security model does this policy describe?","options":["The Zero Trust model, built on verifying explicitly, using least privilege access, and assuming breach","The defense-in-depth model, built on a single hardened perimeter around the network","The shared responsibility model, splitting duties between the consumer and the cloud provider","The single sign-on model, letting one authentication reach many applications"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/7-describe-zero-trust-model","referenceTitle":"Describe Zero Trust model — Microsoft Learn"}}
{"id":"RAN-AZ900-00196","exam":"AZ900","topics":["architecture"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:e05e9c8c1b04787883ab0aab14dca7dd1f7b0508c6265ce1d722f3e65d266d67","aliases":[],"snapshot":{"statement":"A field-services company lets technicians reach work apps from their own phones and home networks. Under the company's new Zero Trust policy, how is a sign-in like this one — from an unmanaged device on a public network — handled?","options":["It is allowed without further checks, because it arrives through the company VPN tunnel from a known network","It is assessed on identity, device, and risk signals, then allowed, challenged for MFA, or blocked accordingly","It bypasses authentication whenever the target application is classified as low-risk, since trusted users are assumed safe","It is blocked outright, because Zero Trust only allows sign-ins from managed devices on the corporate network"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/7-describe-zero-trust-model","referenceTitle":"Describe Zero Trust model — Microsoft Learn"}}
{"id":"RAN-AZ900-00197","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:648869480b62e506e9f9258941bfa8b115015772991b139cb935ddae3060a7b3","aliases":[],"snapshot":{"statement":"A city government's IT architect presents the security design for a new records system: protections are stacked in multiple layers with the records at the center, and if an attacker gets through one layer, the later layers are still in place to slow the attack down and raise alerts.\nWhich security model is the architect describing?","options":["The Zero Trust model","The defense-in-depth model","The shared responsibility model","The consumption-based model"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/8-describe-defense-depth","referenceTitle":"Describe defense-in-depth — Microsoft Learn"}}
{"id":"RAN-AZ900-00198","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:8d82b5ce43af923b3e52f791663094532c80d972869308e3b9bdd9bb5448de4c","aliases":[],"snapshot":{"statement":"While mapping controls for its defense-in-depth strategy, an airline's security team reaches the layer that directly protects customer records and operational data — the information the airline is ultimately trying to safeguard.\nWhich layer of the defense-in-depth model is this?","options":["The application layer","The network layer","The data layer","The perimeter layer"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/8-describe-defense-depth","referenceTitle":"Describe defense-in-depth — Microsoft Learn"}}
{"id":"RAN-AZ900-00199","exam":"AZ900","topics":["architecture"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:b368a35b0236048167162bbc522ff6d1cff0a72e2b173a5715d6b18785a9e0e8","aliases":[],"snapshot":{"statement":"A utility company runs workloads in Azure, in other clouds, and on on-premises servers. Its security team wants one tool that monitors the whole estate, continuously assesses its security posture, and provides recommendations and alerts to improve it.\nWhich service should the team use?","options":["Microsoft Defender for Cloud, for security posture management and workload protection","Azure Monitor, for collecting and analyzing telemetry such as metrics and logs","Azure Advisor, for cost, performance, and reliability recommendations across subscriptions","Microsoft Purview, for data governance and classification across the data estate"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-azure-identity-access-security/9-describe-microsoft-defender-for-cloud","referenceTitle":"Describe Microsoft Defender for Cloud — Microsoft Learn"}}
{"id":"RAN-AZ900-00200","exam":"AZ900","topics":["architecture"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:32bf988c05e140bc4a435297dd16242277061c710ad9254f3f40f0c9721f8fc3","aliases":[],"snapshot":{"statement":"The new CIO of a regional bank keeps mixing up the bank's security tools and asks the security lead for one accurate sentence about Microsoft Defender for Cloud.\nWhich statement is accurate?","options":["It enforces resource tagging rules and allowed virtual machine sizes at deployment time, rejecting any resources that do not match those rules","It collects and analyzes application performance telemetry, such as request rates, failure counts, and page load times","It manages employee identities and single sign-on for the bank's cloud applications, including multifactor authentication","It provides a comprehensive view of security posture across cloud and on-premises resources, combining posture management with workload protection"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/defender-for-cloud/defender-for-cloud-introduction","referenceTitle":"Microsoft Defender for Cloud Overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00201","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:681cdd0aa9616a900089cc80dc1bb49380361213bc0635fbcda21a0fa9d9c824","aliases":[],"snapshot":{"statement":"A software company's monthly Azure bill came in well above forecast. During a cost-review meeting, the team evaluates four candidate actions, and the facts stated in each option have already been confirmed. Which action will NOT reduce the company's Azure bill?","options":["Resize several underutilized virtual machines to a smaller size, lowering their hourly rate","Delete managed disks that are no longer attached to any virtual machine","Remove the descriptive tags the team applies to resources for reporting","Move a development workload to a nearby region where the same virtual machine configuration is offered at a lower price"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/2-describe-factors-affect-costs-azure","referenceTitle":"Describe factors that can affect costs in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00202","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:72224d164352515baa8dffb1fc758c1fd87c3a826971cd35cd87c3b3ac85db27","aliases":[],"snapshot":{"statement":"An online gift retailer adds extra virtual machines and storage for the December shopping season and removes them in January. December's Azure bill is far higher than February's, although the subscription itself never changed. Which characteristic of Azure pricing explains the difference?","options":["Azure follows a consumption-based model: the company pays for the resources used during each billing cycle, so the bill rises and falls with actual usage","Azure recalculates the subscription's tier at the end of each month, and December was billed at a higher tier than February","Azure charges for the peak capacity the subscription could have used during the billing cycle, whether or not that capacity was actually used by any workload","Azure bills a per-deployment fee for each resource created or removed, so the invoice reflects how many deployments were made during the cycle"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/2-describe-factors-affect-costs-azure","referenceTitle":"Describe factors that can affect costs in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00203","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:4646bf87fbeadf73f90f2c3d94cdc064c9c971d2ff30e9993b15b23074bbccf7","aliases":[],"snapshot":{"statement":"A photography platform runs two storage accounts in the same Azure region, each holding a similar volume of photos. One account keeps archives on a cool access tier with limited redundancy; the other serves the galleries from a hot tier with geo-redundant storage. The monthly bills differ noticeably. Which factor explains the difference?","options":["The age of each storage account, because Azure applies legacy rates to accounts that were created before the current pricing release came into effect","The settings chosen for each storage account, such as the access tier and how many redundant copies of the data are kept","The subscription that owns each storage account, because every account in a subscription is billed at one shared rate","The tools the teams use to upload photos, because Azure charges storage accounts different rates depending on the upload tool"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/2-describe-factors-affect-costs-azure","referenceTitle":"Describe factors that can affect costs in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00204","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3b3f45c5d8a45a447f1b2a6863c5de1e9eca4d4cacc2d3a1282f16aeb2f995a5","aliases":[],"snapshot":{"statement":"A video production company is planning a new rendering service and compares identical virtual machine configurations in two Azure regions using Microsoft's estimation tooling. The estimates show different monthly totals for the same configuration. The company asks what makes the same resource cost different amounts in different regions. Which explanation is correct?","options":["The estimate includes a currency conversion fee that Azure adds automatically for regions outside the company's billing country","One region is newer than the other, and Azure temporarily discounts resources provisioned in regions that opened recently","The cost of power, labor, taxes, and fees differs by location, so Azure prices the same resource differently per region","The estimate adds an availability margin that grows with the distance between the chosen region and the location of the company's head office"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/2-describe-factors-affect-costs-azure","referenceTitle":"Describe factors that can affect costs in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00205","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a18fee67fd6fb39ca82cd08bd25636fc1260a193ec907a7409bc1a03641c0ff7","aliases":[],"snapshot":{"statement":"An IT services team decommissioned six virtual machines when a client project ended last quarter, but the following Azure invoices barely decreased. The virtual machines no longer appear anywhere in the subscription. What is the most likely explanation, and what should the team do?","options":["Azure keeps billing deleted virtual machines for one full quarter after deletion, and the charges will stop on their own","The subscription still carries a reservation fee for the deleted virtual machines, so the team should open a support request to remove it","Virtual machine charges are invoiced in advance and reconciled later in a separate process, so the team should expect the bill to drop only within a few months","Deleting a virtual machine does not necessarily delete its attached disks and network interfaces, so the team should locate and remove the still-billed leftover resources"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/2-describe-factors-affect-costs-azure","referenceTitle":"Describe factors that can affect costs in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00206","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:5712bef249bf6bee03948f6cff75ca4bdb1a66a3bb2d16291223469a7f8c3f80","aliases":[],"snapshot":{"statement":"A network engineer at a logistics company is reviewing the data-transfer section of the Azure invoice. Some inbound transfers show no charge, while outbound transfers are grouped by what the invoice calls billing zones. The engineer has worked with availability zones and assumes the two terms refer to the same thing. Which explanation is accurate?","options":["Inbound and outbound transfers are both billed at one flat per-gigabyte rate that depends only on the size of the source virtual machine","Some inbound transfers into Azure datacenters are free, while outbound transfers are priced by billing zones — geographic groupings of regions for data-transfer pricing, distinct from availability zones","Transfers are free while they stay inside a single availability zone, and charges apply only when traffic leaves that availability zone","Outbound transfers between Azure regions are free of charge, and data-transfer billing applies only to traffic that leaves Azure entirely to the public internet"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/2-describe-factors-affect-costs-azure","referenceTitle":"Describe factors that can affect costs in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00207","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:2297f678c9f852a07b5553b6d4bd32867002e7f9ae08424b9d92ed44952e816d","aliases":[],"snapshot":{"statement":"A manufacturer's production database and its reporting virtual machines run continuously, around the clock, with a stable and predictable workload profile. Finance has asked the cloud team to reduce the compute bill without changing the architecture. Which purchasing approach best fits this workload?","options":["Buy a Reservation that commits to the specific resource capacity for a one-year or three-year term and applies discounted pricing to matching usage","Use Spot pricing to consume unused Azure capacity at reduced prices, accepting that the workloads can be evicted when Azure needs the capacity back","Commit to an hourly spend amount across eligible compute services with the Azure savings plan for compute, keeping full flexibility over which services consume it","Purchase the database and reporting software through Azure Marketplace, where third-party vendors set the billing and offer reduced rates for continuous use"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/7a-describe-cost-optimization-options","referenceTitle":"Describe cost optimization options in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00208","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:31130768a28cbd0e6c738413dd7d661fb9b0a8abc14632bc0b7fe4c42c1513cd","aliases":[],"snapshot":{"statement":"An event ticketing platform expects a major sales surge. The engineers debate two ways to prepare: scale out by adding identical virtual machine instances, or scale up by moving to larger virtual machines with more cores and memory. The finance lead asks how each strategy would affect the Azure bill. Which explanation is accurate?","options":["Both strategies raise only the compute line item of the bill; storage and network charges stay flat no matter how many instances run or how large each one is","Scaling out is billed as one flat platform fee per hour regardless of the instance count, so the total only changes when the fleet scales up to larger virtual machines","Scaling up automatically converts the whole deployment to fixed reservation pricing for the commitment term, while scaling out keeps the standard variable per-usage billing","Scaling out multiplies the number of billed instances — each added virtual machine brings its own compute, storage, and network charges — while scaling up moves each instance to a higher-priced size"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/2-describe-factors-affect-costs-azure","referenceTitle":"Describe factors that can affect costs in Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00209","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:005c62899995bb8b45e8a4fcda0e977e01ed81e399f2e04261db7ddf6f77a7b2","aliases":[],"snapshot":{"statement":"A startup is planning its first Azure solution: a web application with a managed database and object storage. Leadership wants a realistic estimate of the monthly cost before any resources are created. Which tool should the team use?","options":["Microsoft Cost Management, after deploying a small pilot of the solution to generate real usage data","Azure Advisor, which forecasts the cost of planned resources before anything is provisioned","The Azure pricing calculator, which estimates the cost of a planned configuration of Azure products","Azure Migrate, which prices new Azure solutions based on an inventory of the company's existing hardware"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/3-compare-pricing-total-cost-of-ownership-calculators","referenceTitle":"Explore the pricing calculator — Microsoft Learn"}}
{"id":"RAN-AZ900-00210","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:0a7b718a4c4ddfcafb4de4beb35c451686c2b5529f27bcd5bf5edf58b6d63a4f","aliases":[],"snapshot":{"statement":"A junior engineer at a logistics software firm is asked to build an estimate for a planned deployment in the Azure pricing calculator. She hesitates, worried that selecting products for the estimate will create resources on the subscription and start charges. What should her teammates tell her?","options":["The pricing calculator is for information purposes only: nothing is provisioned and no services are charged for the items added to an estimate","Products selected in the calculator are provisioned in a stopped state and begin billing only when they are started","Each saved estimate places a one-time estimation fee on the subscription, which is refunded automatically when the estimate is deleted","Estimates act as binding price quotes, so the subscription is charged the estimated amount on the next invoice after the estimate is first saved"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/3-compare-pricing-total-cost-of-ownership-calculators","referenceTitle":"Explore the pricing calculator — Microsoft Learn"}}
{"id":"RAN-AZ900-00211","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:fd3302b6c126b0cde7aa571fb3441a32168cddd63f9947edf2b15c4be1acae94","aliases":[],"snapshot":{"statement":"A streaming media company is designing a new customer portal. Before deploying anything, the architects want to compare the estimated monthly cost of the planned app service tier, managed database, and storage under two different regions and two redundancy options. Which approach meets this requirement?","options":["Deploy a minimal version of the portal in each region and compare the invoices after the first full month","Create a budget for each combination of region and redundancy and compare which budgets raise alerts first during a pilot deployment","Ask Azure Advisor to forecast the cost of each combination of region and redundancy before anything is deployed","Build the planned configuration in the pricing calculator and compare the monthly estimates across the regions and redundancy options"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/3-compare-pricing-total-cost-of-ownership-calculators","referenceTitle":"Explore the pricing calculator — Microsoft Learn"}}
{"id":"RAN-AZ900-00212","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:6b2050aa9d72364645e9775274d7b4420c0dffaae2d73baa4a3c109007efbcce","aliases":[],"snapshot":{"statement":"A conglomerate's CFO asks the IT team to use Microsoft's Total Cost of Ownership (TCO) calculator to build the business case for moving workloads to Azure, following a recommendation in an internal document written a few years ago. When the team checks Microsoft's current documentation, what will they find?","options":["The TCO calculator is now a built-in view of Microsoft Cost Management that compares on-premises invoices with Azure invoices automatically","The TCO calculator has been retired, and the pricing calculator is the current tool for estimating the cost of Azure resources","The TCO calculator was renamed Azure Migrate, which now produces the official cost estimates for migration projects","The TCO calculator is still available, but access is limited to Enterprise Agreement customers with an active prepayment balance"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/3-compare-pricing-total-cost-of-ownership-calculators","referenceTitle":"Explore the pricing calculator — Microsoft Learn"}}
{"id":"RAN-AZ900-00213","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:512ff7a1c1d0f25f98ecedf8d0303742411e43df28943affe920a79946750a61","aliases":[],"snapshot":{"statement":"A healthcare group's Azure spend has grown quickly across several subscriptions. The new finance director wants one place to analyze where the money goes, receive notifications about spending, and set spending limits for the teams. Which capability should the cloud team present first?","options":["Microsoft Cost Management, a set of tools to analyze, monitor, and manage Azure costs — including cost reporting, budgets, and alerts","Azure Advisor, which analyzes resources that are already deployed and enforces spending limits on underutilized services","The Azure pricing calculator, which tracks the realized monthly costs of all of the subscriptions in a single consolidated view","Azure Service Health, which sends billing notifications to subscription owners whenever service prices change in their regions"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/6-describe-azure-tool","referenceTitle":"Describe the Microsoft Cost Management tool — Microsoft Learn"}}
{"id":"RAN-AZ900-00214","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:99b675184f758ea0d28ba1650f99190e262d016c97cc640d3186184420cf9550","aliases":[],"snapshot":{"statement":"A fintech's operations analyst must investigate last month's Azure bill: which services drove the costs, how they accumulated over the month, and whether spending is trending above the team's target. Which Cost Management feature should the analyst open first?","options":["Budgets, which visualize historical invoices and automatically forecast the monthly costs of the next billing period for every team in the account","Cost alerts, which display each service's accumulated costs in a daily email digest sent to the subscription owners","Cost analysis, which breaks down costs by subscription, resource group, or service and shows accumulated costs over time against a budget","The pricing calculator, which can reconstruct last month's deployed configuration and calculate the charges it actually incurred"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/quick-acm-cost-analysis","referenceTitle":"Quickstart: Start using Cost Analysis — Microsoft Learn"}}
{"id":"RAN-AZ900-00215","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7f421a9efb594373156187c3be95a4f8f3ed251f309a6551962b4c0c84ac5d6d","aliases":[],"snapshot":{"statement":"A game studio's nonproduction subscription keeps burning through its monthly allowance without anyone noticing until the invoice arrives. The studio wants spending tracked against a monthly target, and, when the target's threshold is crossed, the affected dev-test resources shut down automatically. Which configuration meets this requirement?","options":["A cost analysis view scheduled weekly, which the team reviews to decide manually which resources to stop before the next invoice","A spending cap that Azure Advisor attaches to each individual resource when it detects low utilization, stopping idle resources automatically at the end of each month","A budget on its own, because reaching a budget automatically stops all resource consumption in its scope","A budget with an alert threshold, plus the budget's automation configured to suspend or modify the target resources when the threshold is reached"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/6-describe-azure-tool","referenceTitle":"Describe the Microsoft Cost Management tool — Microsoft Learn"}}
{"id":"RAN-AZ900-00216","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:005171ad996d19115fb693f90cf90297415ddd262f528cad7d411fe329834093","aliases":[],"snapshot":{"statement":"The research computing lead at a university wants an email sent to the lab's shared mailbox as soon as monthly spending on a production resource group reaches 80% of the agreed amount. Which setup meets this requirement?","options":["A budget on the resource group with an alert threshold at 80%, generating a budget alert automatically and emailing the configured recipients when spending reaches it","An Azure Monitor metric alert that watches the resource group's spending and emails the shared mailbox when the metric crosses the 80% mark","A department spending quota alert, which emails the department owners when their spending quota for the period is consumed","A scheduled export of the resource group's cost analysis chart, emailed to the shared mailbox on the first day of every month"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/tutorial-acm-create-budgets","referenceTitle":"Tutorial: Create and manage budgets — Microsoft Learn"}}
{"id":"RAN-AZ900-00217","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:fa68b68b96d97e08aa22c6ef98409bf682164b6407d3bca81c0723a9ea712608","aliases":[],"snapshot":{"statement":"An automotive supplier with an Enterprise Agreement funds its Azure usage with Azure Prepayment, previously called monetary commitment. The billing administrator wants alerts that fire as the prepayment balance is drawn down. Which Cost Management alert type serves this purpose, and how does it behave?","options":["Budget alerts, which the billing team must generate manually from the cost alerts page on a weekly schedule","Credit alerts, which are generated automatically at 90% and 100% of the Azure credit balance","Department spending quota alerts, which fire automatically when any subscription in the enrollment exceeds its assigned spending quota","Invoice alerts, which are raised when the monthly invoice consumes the remaining prepayment balance on the account"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/6-describe-azure-tool","referenceTitle":"Describe the Microsoft Cost Management tool — Microsoft Learn"}}
{"id":"RAN-AZ900-00218","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:45accfe7b4714666a669273f73e3caaa6d00107955cdaa85a7c0edbb48beb0fb","aliases":[],"snapshot":{"statement":"After a busy quarter, a marketing agency's Azure bill is far above expectations. The environment includes many development and test virtual machines that are suspected to sit idle most of the time. The team wants personalized recommendations that identify underutilized resources and suggest concrete ways to reduce spend. Which capability should they use?","options":["The pricing calculator, which flags resources in the current configuration whose estimated cost is above the account average","Budgets, which recommend which virtual machines should be shut down once the monthly spending threshold is crossed","Azure Advisor cost recommendations, which analyze resource utilization and recommend, for example, shutting down or resizing low-utilization virtual machines","Cost analysis, which automatically proposes configuration changes whenever the costs inside a resource group rise above their recent trend"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/advisor/advisor-cost-recommendations","referenceTitle":"Optimize virtual machine (VM) or virtual machine scale set (VMSS) spend by resizing or shutting down underutilized instances — Microsoft Learn"}}
{"id":"RAN-AZ900-00219","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ee589d7862e9b2bc40e3e44387fe0b42032390f4d82fa69077858d0f6b48f21a","aliases":[],"snapshot":{"statement":"A holding company runs workloads for two independent business units in a single Azure subscription. Each unit must now receive its own Azure invoice, and each unit's costs must be reported separately. What should the company do?","options":["Keep the single subscription and attach a business-unit tag to every resource, so the invoices are split by tag at the end of each month","Create a separate resource group for each business unit, because Azure issues one invoice per resource group in a subscription","Create a management group for each business unit, because Azure issues one invoice per management group in the organization","Create a separate subscription for each business unit, so each unit's usage is billed and reported at its own subscription scope"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/cost-management-billing/cost-management-billing-overview","referenceTitle":"Overview of Billing — Microsoft Learn"}}
{"id":"RAN-AZ900-00220","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:285b041b1a60d9a4393eb70e49e2f706c72939a48579935f88777dac73095189","aliases":[],"snapshot":{"statement":"A fast-growing SaaS provider wants a lightweight way to attach structured information — such as environment, owner, and cost center — to its Azure resources so they can be organized, searched, and reported on. Which statement describes the construct the team should use?","options":["Resource locks, which label resources with metadata and prevent the labels from being changed by other administrators","Resource tags, which are name-value pairs that add metadata to resources for organization and reporting","Management groups, which attach key-value metadata to every subscription that sits beneath them in the hierarchy","Resource groups, which store a fixed set of metadata fields for each resource they contain when it is created"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/7-describe-purpose-of-tags","referenceTitle":"Describe the purpose of tags — Microsoft Learn"}}
{"id":"RAN-AZ900-00221","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:dc7145fc8526835e761cd612b9312f6f708034f552048c04816e9b163777e536","aliases":[],"snapshot":{"statement":"An engineering consultancy keeps all of its Azure resources in one subscription, with resource groups organized by application. Finance now needs monthly cost reports by internal cost center, a categorization that cuts across applications, without reorganizing how the applications are grouped. How should the consultancy meet this requirement?","options":["Apply a cost center tag to each resource and group costs by that tag in cost reporting","Move the resources into new resource groups so that each resource group maps to a single cost center","Create a new subscription for each cost center and migrate the resources accordingly","Use Azure Policy to assign each resource to a cost center at the subscription level"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/7-describe-purpose-of-tags","referenceTitle":"Describe the purpose of tags — Microsoft Learn"}}
{"id":"RAN-AZ900-00222","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e90386c963a808733993769b5aa3ef29de9dab87675ff4be36633ddd6878d48d","aliases":[],"snapshot":{"statement":"A streaming provider's platform team runs a maintenance pipeline that must execute a task against every resource belonging to one specific workload. The resources of that workload are spread across multiple resource groups and subscriptions. Which approach lets the pipeline select exactly those resources reliably?","options":["Match resources by name prefix, because every resource of the workload is expected to follow the team's naming convention","Keep a list of the resource groups that contain workload resources and treat group membership as the selector","Apply resource locks to the workload's resources so the pipeline can find them by their lock status","Apply a workload tag to the resources and have the pipeline query for, and act on, the resources carrying that tag"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/7-describe-purpose-of-tags","referenceTitle":"Describe the purpose of tags — Microsoft Learn"}}
{"id":"RAN-AZ900-00223","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:c19b16e9b91ca1e7bc80b3b0a75976b4b43f3b7b4396d9a720c1f3e9070a20a9","aliases":[],"snapshot":{"statement":"An insurer's cloud governance team publishes a tagging standard: every new resource must carry an owner tag, and tags that are accidentally removed should be restored. Manual reviews keep missing violations. Which mechanism should the team use to enforce the standard?","options":["Apply the owner tag to each resource group and rely on the resources inheriting it automatically","Put a resource lock on every resource so that its tags can no longer be removed","Use Azure Policy to require the owner tag on provisioned resources and to reapply tags that have been removed","Export the full resource inventory every week and have a script email the owners of resources that are missing the tag"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-cost-management-azure/7-describe-purpose-of-tags","referenceTitle":"Describe the purpose of tags — Microsoft Learn"}}
{"id":"RAN-AZ900-00224","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:52656278aa3c1cf6d5a6a2fe13819d8db9ebce530d5c54b20cd7ec7e48269d21","aliases":[],"snapshot":{"statement":"A retail chain's architect applies the tag CostCenter=100 to a resource group that contains several virtual machines. A later inventory report that lists each virtual machine's own tags shows none of them carrying the cost center. What is the most likely explanation?","options":["The tag will appear on the virtual machines within 24 hours, once Azure's tag synchronization job completes its daily run","Tags applied to a resource group are not automatically applied to the resources inside it — each level carries its own tags","The virtual machines reject the tag because their current SKU does not support cost center metadata on this resource type","The tag was applied in a different region from the virtual machines, so it cannot reach any of the resources outside of its own region"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/tag-resources","referenceTitle":"Use tags to organize your Azure resources and management hierarchy — Microsoft Learn"}}
{"id":"RAN-AZ900-00225","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:ea0519c77c9e829f55de18ae76fd642694b00407a9613038cfb30a5973d20e69","aliases":[],"snapshot":{"statement":"A pharmaceutical company on an Enterprise Agreement enables the tag inheritance setting in Microsoft Cost Management. Its resource groups carry cost center tags, but most child resources have no tags of their own. Which statement describes what changes once the setting takes effect?","options":["Subscription and resource group tags are applied to child resources' usage records in cost data, enabling grouping by tag while resources stay untagged","Every child resource immediately receives the inherited tags on its own properties, including resources that were created years before the setting was enabled","Only resources created after the setting is enabled inherit the resource group tags, while existing resources stay excluded from any grouping","Inherited tags override any tags already applied directly to a resource in every case where the two values differ from each other"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/enable-tag-inheritance","referenceTitle":"Group and allocate costs using tag inheritance — Microsoft Learn"}}
{"id":"RAN-AZ900-00226","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:a78dd641920dffa6a5ea7e0fe19df70be78c5985134920e1de9f3111527d02cb","aliases":[],"snapshot":{"statement":"A national hospital network keeps clinical and operational data in several places: an on-premises SQL Server instance, Azure data services, and a third-party software-as-a-service analytics platform. The newly hired chief data officer wants a single solution that provides one unified view of all this data and helps govern it. Which service should the chief data officer choose?","options":["Microsoft Purview, which provides a single, unified view into data across on-premises, multicloud, and software-as-a-service environments","Azure Arc, which extends Azure resource management to servers and clusters that run outside of Azure","Azure Monitor, which collects and analyzes telemetry such as metrics and logs from applications and infrastructure","Azure Policy, which enforces organizational standards and evaluates the compliance of Azure resources"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/2-describe-purpose-microsoft-purview","referenceTitle":"Describe the purpose of Microsoft Purview — Microsoft Learn"}}
{"id":"RAN-AZ900-00227","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:50443afaabf5523c97bbf5932f0d0b21626b2bc0d1750fa5ea509ee1a8210f8d","aliases":[],"snapshot":{"statement":"Analysts at an online retailer waste hours every week asking engineers where datasets live and what each column means. The data team wants a governed, searchable experience where business users can discover available data assets, understand what they need to know about them, and get secure access. Which solution should the team choose?","options":["The catalog experience in Microsoft Purview, which helps data consumers find, understand, and access data securely","Azure Data Factory, which builds pipelines that copy and transform data between sources and destinations","Azure Data Share, which distributes data assets to external partners and other outside organizations","Azure Synapse Analytics, which consolidates data from across the company into a single warehouse for large-scale reporting"],"answer":"A","reference":"https://learn.microsoft.com/en-us/purview/unified-catalog","referenceTitle":"Learn about Microsoft Purview Unified Catalog — Microsoft Learn"}}
{"id":"RAN-AZ900-00228","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:7e075a50792e4e678aeb953b4cc07d85863a3a49d82b389ae28c41622ff082c9","aliases":[],"snapshot":{"statement":"An external auditor is questioning the figures in a logistics company's quarterly fuel-cost report and asks exactly how each number was derived. The data platform team must show the data's full journey — from the source systems through every transformation to the final report. Which Microsoft Purview capability should the team use?","options":["Data Lifecycle Management, which retains and deletes records according to a defined retention schedule","Data Loss Prevention, which blocks sensitive data from leaving the organization through sharing","End-to-end data lineage, which traces how data flows from sources through every transformation to consumers","Sensitive data classification, which labels data assets that contain regulated information"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/2-describe-purpose-microsoft-purview","referenceTitle":"Describe the purpose of Microsoft Purview — Microsoft Learn"}}
{"id":"RAN-AZ900-00229","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:42db4e3b68e7dc27c044c4fb5514db979910c532a9acd75fe5ec29cf968d2ff6","aliases":[],"snapshot":{"statement":"A compliance officer at an insurance group must build an inventory of where sensitive information — such as policyholders' identification numbers and health details — is stored across databases, file shares, and on-premises systems. She wants automated scanning that discovers this data and applies classification across the whole estate. Which service meets this requirement?","options":["Microsoft Defender for Cloud, which assesses the security posture of cloud workloads and issues recommendations","Azure Migrate, which discovers on-premises servers and assesses their readiness for migration","Azure Policy, which audits resource configurations against organizational standards","Microsoft Purview, which scans data sources across the estate and applies sensitive data classification"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/2-describe-purpose-microsoft-purview","referenceTitle":"Describe the purpose of Microsoft Purview — Microsoft Learn"}}
{"id":"RAN-AZ900-00230","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:453d25b81baee33623d656479cf29105f77093ecc6331f0515a670d247054686","aliases":[],"snapshot":{"statement":"A streaming media company is defining its governance program. The platform team must enforce rules on Azure resources, such as deploying only to approved regions. Separately, the data office must discover, catalog, and classify company data wherever it lives — in Azure, in other clouds, and on-premises. Which pairing of services meets both requirements?","options":["Microsoft Purview to enforce the resource rules, and Azure Policy to discover, catalog, and classify the data","Azure Policy to enforce the resource rules, and Microsoft Purview to discover, catalog, and classify the data","Azure Policy for both requirements, because policy definitions can evaluate the data stored inside resources","Microsoft Purview for both requirements, because its data map can deny the creation of noncompliant resources"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/2-describe-purpose-microsoft-purview","referenceTitle":"Describe the purpose of Microsoft Purview — Microsoft Learn"}}
{"id":"RAN-AZ900-00231","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:89f53ba8b65cea26503ca7584fa222069eb1b4a7d4f5933deceeacf339ee80a1","aliases":[],"snapshot":{"statement":"During onboarding week at an energy utility, a newly hired cloud engineer is asked to summarize Microsoft Purview for the rest of the team. Several teammates currently confuse it with other governance services in Azure. Which statement about Microsoft Purview is accurate?","options":["Microsoft Purview enforces organizational standards on Azure resources and can prevent noncompliant resources from being created","Microsoft Purview locks critical resources so that not even administrators can update or delete them","Microsoft Purview builds a unified data map of the organization's entire data estate, spanning on-premises, multicloud, and software-as-a-service sources","Microsoft Purview forecasts cloud spending by subscription and raises alerts when budgets are exceeded"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/2-describe-purpose-microsoft-purview","referenceTitle":"Describe the purpose of Microsoft Purview — Microsoft Learn"}}
{"id":"RAN-AZ900-00232","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6daba9a8a8e426f994440e2af16ee7daeab1b07e139d392b0feeded63f5c1d2f","aliases":[],"snapshot":{"statement":"The IT operations lead at a university must ensure that resources across several Azure subscriptions follow institutional standards — for example, only approved virtual machine sizes and required tags — and must be able to show auditors which resources comply. Which service is designed for this need?","options":["Azure Policy, which evaluates resources against your rules and reports each resource's compliance state","Azure role-based access control (RBAC), which controls which users can perform actions on each resource","Resource locks, which prevent resources from being deleted or changed","Azure Monitor, which collects metrics and activity logs for operational insight"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/3-describe-purpose-azure-policy","referenceTitle":"Describe the purpose of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00233","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0492f55e01b7b91b59eb4777f34e10ab3f0ff789f27254c0178a6fcc4cb917ca","aliases":[],"snapshot":{"statement":"A fintech's platform team has drafted a policy that requires every storage account to carry an Owner tag. Before enforcing the rule, they want to track how it would affect existing and new resources without blocking anyone's deployments. Which policy effect should they start with?","options":["Deny, so that any request missing the tag is rejected","Modify, so that the missing tag is added to every request automatically","Audit, so that noncompliant resources are recorded in the compliance results without being blocked","Disabled, so that the definition exists in the environment but is never evaluated"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/governance/policy/overview","referenceTitle":"Overview of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00234","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a1ccb41d8677bef9c84ee8069bd26e781ef94156c0325b14e1866109770f55cd","aliases":[],"snapshot":{"statement":"The security team at a manufacturing company must guarantee that no new virtual machine can be created outside the company's approved Azure regions — an attempt to create a VM in another region has to fail. Which effect should the team's policy assignment use?","options":["Audit, which lets the VM be created and then flags it as noncompliant","Deny, which blocks the creation request so the noncompliant VM is never created","Modify, which changes the requested region to an approved location before the VM is created","AuditIfNotExists, which checks the VM after deployment and logs a noncompliance record"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/governance/policy/overview","referenceTitle":"Overview of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00235","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a0499fa873f8e3d63d07af9dbed7ced9867a551740e473e30ac017872164e4f5","aliases":[],"snapshot":{"statement":"A cloud architect at a public-sector agency needs to roll out a security baseline made of many related policy definitions — allowed locations, required tags, and allowed virtual machine SKUs. The definitions must be assigned and tracked together as a single item, and more definitions may be added later without creating extra assignments. What should the architect create?","options":["A management group, which bundles the policies for all subscriptions beneath it","One large custom policy definition that contains all of the individual rules","A separate assignment for each policy definition, so each can be tracked on its own","An initiative definition, which groups the policy definitions for assignment and tracking as a set"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/governance/policy/overview","referenceTitle":"Overview of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00236","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f11a6fc352c73a1193a1d43fcbd628378e20800c63862fc07b6925346cee9fe4","aliases":[],"snapshot":{"statement":"A holding company organizes its Azure environment under a management group that contains multiple subscriptions. Governance staff want a single policy assignment that covers every subscription and all resources inside them — including resources created in the future. Where should the assignment be created?","options":["At the management group scope, because all child subscriptions and their resources inherit the assignment","At each subscription separately, because policy assignments are never inherited across scopes","At each resource group, because policies apply only within the resource group where they're assigned","At the individual resource scope, because Azure Policy evaluates resources one by one"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/governance/policy/overview","referenceTitle":"Overview of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00237","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:0194dc1782d588ba10ea64421822798c9fae02eff4ad23e15de8f860953fcce9","aliases":[],"snapshot":{"statement":"An administrator at a food delivery company has assigned a policy that requires an Environment tag on resources. The director now asks for a list of the resources that are missing the tag. Where should the administrator look for this information?","options":["Azure Advisor, which recommends best practices for cost, performance, reliability, and security","Azure Service Health, which lists Azure service incidents and planned maintenance","Azure Monitor activity logs, which record the operations performed on each resource over time","The compliance results in Azure Policy, which show each resource as compliant or noncompliant"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/governance/policy/overview","referenceTitle":"Overview of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00238","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:960b57a2bef332eaad192fb7ec8be56db24c6ff9205e9c3fa6c192ea637f4326","aliases":[],"snapshot":{"statement":"A software company requires every new Azure resource to carry a CostCenter tag. Deployments that omit the tag must not be rejected — instead, Azure should add the tag with a default value during the create or update request. Which effect should the policy definition use?","options":["Audit, which records the resource as noncompliant in the results and leaves the tag missing","Modify, which adds the required tag with its default value when the request omits it","Deny, which rejects any request that doesn't already include the tag","DeployIfNotExists, which deploys a related resource when the tag is absent"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/governance/policy/overview","referenceTitle":"Overview of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00239","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:6c2c07a290acd7bc044f62d0e6a77b48ac482a52e7a5b879793996a36a67e0b1","aliases":[],"snapshot":{"statement":"A junior cloud administrator at a startup asks a senior engineer to clarify the difference between Azure RBAC and Azure Policy. The senior engineer explains that Azure RBAC controls which actions each user can perform. According to that explanation, what does Azure Policy control?","options":["Which users are allowed to read the data that is stored inside each Azure resource","The properties resources can have, such as locations, SKUs, and tags, regardless of who creates them","Which users are allowed to delete each resource in the subscription","The prices and budget limits that apply to each resource"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/governance/policy/overview","referenceTitle":"Overview of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00240","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:855559c452a3bd0da58aa4f2dc5c6521f747e6a8b99cd006620310a93cf467d3","aliases":[],"snapshot":{"statement":"A subscription-level policy assignment blocks the creation of networking resources everywhere in a research organization's subscription. A new dedicated resource group must be exempt from this rule so the networking team can create resources there, while the policy continues to apply everywhere else. What should the administrator do?","options":["Delete the assignment and recreate it on every other resource group individually","Grant the networking team Owner permissions on the new resource group, which bypasses the policy for them","Update the assignment so the new resource group is excluded from the policy's scope","Apply a CanNotDelete lock to the new resource group, which stops policy evaluation inside it"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/governance/policy/overview","referenceTitle":"Overview of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00241","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:a5a536a37d25f050ffd13f4f776da5dcb5dd22b61c58c77d3bd4d1859b0a66b0","aliases":[],"snapshot":{"statement":"An airline's subscription contains hundreds of virtual machines, several of which use unapproved VM sizes. An administrator assigns a new Audit-effect policy that permits only approved sizes, and assumes it will affect only virtual machines created from now on. What actually happens to the existing noncompliant VMs?","options":["They're never evaluated, because Azure Policy applies only to resources created after the assignment exists","They're automatically resized to an approved SKU, because Audit remediates noncompliant resources","The assignment fails, because the subscription already contains noncompliant resources","They're evaluated and flagged as noncompliant in the policy's compliance results"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/3-describe-purpose-azure-policy","referenceTitle":"Describe the purpose of Azure Policy — Microsoft Learn"}}
{"id":"RAN-AZ900-00242","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:f8a3b9fe2262986b9ba1663ab3560a254049115be660e27d556b0b7722c27598","aliases":[],"snapshot":{"statement":"During a cleanup sprint at a utilities company, an operator deleted the wrong resource group and caused a production outage. Management asks for the simplest built-in guardrail that prevents anyone from accidentally deleting the critical production database, while normal configuration changes must continue. What should the team apply?","options":["A ReadOnly lock, which blocks both updates and deletions of the database","A Delete (CanNotDelete) lock, which lets authorized users modify the database but blocks its deletion","A tag named Protection with the value DoNotDelete, which marks the database as critical","An Azure Policy assignment with the Audit effect, which reports any attempt to delete the database"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/4-describe-purpose-resource-locks","referenceTitle":"Describe the purpose of resource locks — Microsoft Learn"}}
{"id":"RAN-AZ900-00243","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:d86e1c84e18aaca3fd14cf0f5caa600e5aeab9726db851ca2207d25c2b3c9e57","aliases":[],"snapshot":{"statement":"A pharmacy chain's production database server has a Delete (CanNotDelete) lock applied. An authorized database engineer needs to apply a routine configuration update to the server. What happens when the engineer submits the change?","options":["The update succeeds, because a Delete lock still allows authorized users to modify the resource","The update fails, because every resource lock blocks all write operations on the resource","The update succeeds only after the engineer asks the subscription Owner to remove the lock","The update fails, because a Delete lock freezes the resource configuration until the lock expires"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/4-describe-purpose-resource-locks","referenceTitle":"Describe the purpose of resource locks — Microsoft Learn"}}
{"id":"RAN-AZ900-00244","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:82fe6ff33509f83d35eed85f40560905dc19d5239f84076e4719052f84ef62a0","aliases":[],"snapshot":{"statement":"An external audit at a pension fund requires that, for the next 60 days, nobody may change or delete the resources used by the claims system in any way, though auditors must still be able to view them. Which lock level meets this requirement?","options":["CanNotDelete, because it blocks changes and deletions while still allowing reads","CanNotDelete, because it blocks deletions while still allowing configuration changes","ReadOnly, because it allows reads but blocks both updates and deletions","ReadOnly, because it allows authorized administrators to keep updating the resources"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/4-describe-purpose-resource-locks","referenceTitle":"Describe the purpose of resource locks — Microsoft Learn"}}
{"id":"RAN-AZ900-00245","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e1c52f64cd320b00fce04ca528e1ac964bdf8c92a0452b829c91027b679135ba","aliases":[],"snapshot":{"statement":"A media company wants to protect all resources in its production subscription from accidental deletion with a single lock, while day-to-day configuration changes must keep working. New resources are added to the subscription constantly. What should the company do?","options":["Apply a Delete (CanNotDelete) lock at the subscription scope, so all existing and future resources inherit it","Apply a Delete lock to each resource individually, because locks affect only the resource they're created on","Apply a Delete lock at the management group above the subscription, because management groups are the highest lockable scope","Apply a ReadOnly lock at the subscription scope, so the resources stay protected while updates continue"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/lock-resources","referenceTitle":"Lock your Azure resources to protect your infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00246","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:77308c2d57867080b82f3da112415db5d47178e57732321b75cc7802ded1c49d","aliases":[],"snapshot":{"statement":"At a mining company, a subscription Owner with full Azure RBAC permissions tries to delete a storage account that has a CanNotDelete lock, and the operation fails. A colleague insists the lock must be misconfigured, because Owner permissions should override it. Which explanation is correct?","options":["The lock is misconfigured, because Owner permissions always take precedence over resource locks","The storage account is being blocked by Azure Policy, so the Owner needs a policy exemption","Owners can delete locked resources from PowerShell or the Azure CLI, just not from the portal","Resource locks apply regardless of RBAC permissions — even an Owner must remove the lock before deleting"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/4-describe-purpose-resource-locks","referenceTitle":"Describe the purpose of resource locks — Microsoft Learn"}}
{"id":"RAN-AZ900-00247","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:eb5ac8a27c992eb9647c69bfca7ee3d60b90ebbce66d586b443b275db08a910b","aliases":[],"snapshot":{"statement":"To reduce accidental deletions, a team lead at an e-commerce company proposes applying a tag named Lock with the value DoNotDelete to every production resource, expecting Azure to block deletions automatically. What is the best response to this proposal?","options":["The tag will work, provided it's applied by a user with Owner permissions","The tag won't block anything — tags are metadata for organizing and reporting, so a Delete lock should be applied instead","The tag will work as long as child resources inherit it from the resource group","The tag won't block anything, but a policy with the Modify effect could enforce it as a deletion guard"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/tag-resources","referenceTitle":"Use tags to organize your Azure resources and management hierarchy — Microsoft Learn"}}
{"id":"RAN-AZ900-00248","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:89edbcdfba07b86395501aa3faa55d6af35168b7a3bdf6ae137e00e7a7f15cbf","aliases":[],"snapshot":{"statement":"An administrator at a bank applies a ReadOnly lock to a production storage account. An automation tool that previously listed the account's access keys to connect now fails on that step, even though listing keys looks like a read-only action. Why did the lock block this operation?","options":["ReadOnly locks on storage accounts also block all data operations on blobs and queues","The tool's service principal no longer has the Reader role on the storage account","The List Keys operation is a POST request, and a ReadOnly lock blocks requests that aren't read-type","The lock disabled the account keys, so only Microsoft Entra credentials work from now on"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/lock-resources","referenceTitle":"Lock your Azure resources to protect your infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00249","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:8dd42c73c2b5324bac0e85f646fc55f3ca472451e520015ea246f0a4c195f4f9","aliases":[],"snapshot":{"statement":"A database administrator applies a ReadOnly lock to the logical server that hosts a critical Azure SQL database and tells management the data is now protected against loss. Management asks whether the lock also prevents data inside the database from being modified or deleted. What is the correct answer?","options":["No — locks apply to control-plane operations, so creating, updating, and deleting data inside the database is still possible","Yes — a ReadOnly lock blocks every operation on the server, including data changes","Yes — locks protect both the resource and the data it contains, so no backup is needed","No — but applying a CanNotDelete lock instead would protect the data while allowing configuration changes"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/lock-resources","referenceTitle":"Lock your Azure resources to protect your infrastructure — Microsoft Learn"}}
{"id":"RAN-AZ900-00250","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:623f2bee542307e0f449f05cab6b821050b291d4f9fd6aaa1a33e2b8adc52789","aliases":[],"snapshot":{"statement":"A payment-processing company defines two requirements: (1) the production payments database must be protected against accidental deletion by anyone — including administrators — while administrators keep their current permissions and routine configuration changes continue; and (2) any new storage account must be created only in approved regions. Which combination meets both requirements?","options":["A ReadOnly lock on the database, and an Azure Policy assignment with the Deny effect for allowed locations","A Delete (CanNotDelete) lock on the database, and an Azure Policy assignment with the Audit effect for allowed locations","A DoNotDelete tag on the database, and an Azure RBAC role that removes delete permissions from administrators","A Delete (CanNotDelete) lock on the database, and an Azure Policy assignment with the Deny effect for allowed locations"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-azure-for-governance-compliance/4-describe-purpose-resource-locks","referenceTitle":"Describe the purpose of resource locks — Microsoft Learn"}}
{"id":"RAN-AZ900-00251","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6a5a564f53ec5a6dda4c89517ae3d654d7a791bd444c862e28e8f99a0182dd83","aliases":[],"snapshot":{"statement":"A regional bank has opened its first Azure subscription. The IT operations team includes administrators with no scripting background who prefer working visually. During onboarding, they ask what the Azure portal offers them.\n\nWhich statement describes the Azure portal?","options":["A desktop application that administrators must install on their workstations before they can manage Azure resources","A browser-based shell where administrators run Bash or PowerShell commands to manage Azure resources","A web-based, unified console that provides a graphical interface for building, managing, and monitoring Azure resources","A command-line executable that automates deployments by processing JSON definition files"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-portal/azure-portal-overview","referenceTitle":"What is the Azure portal? — Microsoft Learn"}}
{"id":"RAN-AZ900-00252","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:c6bbfb8528802d6b6e790f143dbe007b7da21c6b55a885bd608d1c00cb1e32ae","aliases":[],"snapshot":{"statement":"A data analyst at an online education company starts every workday in the Azure portal by opening the same handful of resources: a web app, its database, and two virtual machines. She wants all of them to appear together as soon as she signs in.\n\nWhat should she do in the Azure portal?","options":["Customize a dashboard with the resources she monitors and set that dashboard as her startup page","Place the resources in a management group and open that group's page each morning","Bookmark each resource's page in her web browser and open the bookmarks one by one","Export the resources into an Azure PowerShell script that lists them at the start of each day"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-portal/azure-portal-overview","referenceTitle":"What is the Azure portal? — Microsoft Learn"}}
{"id":"RAN-AZ900-00253","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:fa5e6ef9567ff25eda97c8018d7aaa4e1df30fb8862942c646ea42cd8e60eff7","aliases":[],"snapshot":{"statement":"A small design agency is deciding which Azure management tool fits each kind of task. The agency owner is comfortable with visual tools and wants to understand where the Azure portal is the right choice.\n\nWhich task is the best fit for the Azure portal?","options":["Running an unattended nightly job that creates and then deletes dozens of temporary test resources","Exploring a new Azure service and creating a single storage account with the help of guided wizards","Deploying identical environments across many different subscriptions from one repeatable definition file","Governing on-premises servers in the agency's own datacenter as if they were Azure resources"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-portal/azure-portal-overview","referenceTitle":"What is the Azure portal? — Microsoft Learn"}}
{"id":"RAN-AZ900-00254","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:e34bf479703d97375b11c5011245355bdec79a9f0a1afdb2245c3f0384198594","aliases":[],"snapshot":{"statement":"During onboarding at a logistics company, a new operations engineer asks how the team runs Azure commands without installing anything on their laptops.\n\nWhich statement describes Azure Cloud Shell?","options":["A desktop application that gives each administrator a graphical copy of the Azure console on their workstation","A command-line tool that can only be installed on computers that run a Linux distribution","A JSON file format used to declare the resources that a deployment will create in Azure","An interactive, authenticated, browser-accessible terminal for managing Azure resources, offering a choice of Bash or PowerShell"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/cloud-shell/overview","referenceTitle":"What is Azure Cloud Shell? — Microsoft Learn"}}
{"id":"RAN-AZ900-00255","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:27deefec5feedc44a88677fe248f345f1673cbc6bc5b26b5a55d34fa96776a72","aliases":[],"snapshot":{"statement":"The platform engineers at a video streaming startup are fluent in Bash scripting on Linux but have never used PowerShell. They want a command-line tool they can run from their own terminals to manage Azure resources.\n\nWhich tool is the best fit for the team?","options":["Azure PowerShell, because it can automate management tasks that the Azure CLI cannot","Azure CLI, because its Bash-based command syntax matches the team's existing scripting experience","Azure Cloud Shell, because it is the only environment where the Azure CLI can run","Azure PowerShell, because the Azure CLI cannot be installed on Linux workstations"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-manage-deploy-azure-resources/2-describe-interacting-azure/","referenceTitle":"Describe tools for interacting with Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00256","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:592bd6816d222fad416568a2d2f1cf9562c38e5a87eecfcf68ac184c544e95b8","aliases":[],"snapshot":{"statement":"An insurance company is writing an internal quick-reference guide about its Azure management tools. The documentation owner asks for one accurate sentence about Azure PowerShell.\n\nWhich statement should the guide include?","options":["A product consisting of official PowerShell modules whose cmdlets manage Azure resources by calling the Azure REST APIs","A Bash-based command-line tool that can only be installed on computers running a Linux distribution","A graphical console in the Azure portal that records the manual steps of each configuration change","A browser-accessible terminal that runs cmdlets but cannot be installed on local machines"],"answer":"A","reference":"https://learn.microsoft.com/en-us/powershell/azure/what-is-azure-powershell","referenceTitle":"What is Azure PowerShell — Microsoft Learn"}}
{"id":"RAN-AZ900-00257","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:751df2cbcb00a6bc0233cedcfddf920c06c11c282d24813ce42201ebeb37f853","aliases":[],"snapshot":{"statement":"An auditor at a public-sector agency occasionally runs a few Azure PowerShell cmdlets from different shared workstations. She wants a shell that is already signed in with her Azure account and permissions, requires no installation, and asks for no separate login step.\n\nWhich tool meets these requirements?","options":["Install the Az PowerShell module on each workstation and run a sign-in command before every session","Use the Azure portal's guided wizards to perform equivalent changes without running cmdlets","Use Azure Cloud Shell, which authenticates automatically with the Azure credentials of the signed-in user","Use Azure Arc to project her workstations into Azure and run cmdlets on them from there"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/cloud-shell/overview","referenceTitle":"What is Azure Cloud Shell? — Microsoft Learn"}}
{"id":"RAN-AZ900-00258","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:cffaa2ddde540fadd5ab244f7a62323841421c0fefde5c2e15c3eef405eb4693","aliases":[],"snapshot":{"statement":"An energy company is preparing an internal comparison of Azure command-line tools, and the author needs an accurate definition of the Azure CLI.\n\nWhich statement accurately describes the Azure CLI?","options":["A Windows-only executable that reproduces the Azure portal's guided wizards in a console window","A PowerShell module whose cmdlets can only run inside Windows PowerShell on Windows machines","An interactive terminal that requires an Azure Files share before any command can run","A cross-platform command-line tool for connecting to Azure and executing administrative commands, interactively or in scripts"],"answer":"D","reference":"https://learn.microsoft.com/en-us/cli/azure/what-is-azure-cli","referenceTitle":"What is the Azure CLI? — Microsoft Learn"}}
{"id":"RAN-AZ900-00259","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:fa419d60b65d3d77221277f7ef2c81681fb6ec3ea3fa72e7b2f886531b3da1a0","aliases":[],"snapshot":{"statement":"Two engineers at a manufacturing firm are comparing Azure command-line tools before the team standardizes on one. One argues that the tools serve different workload sizes; the other disagrees and wants a definitive statement for the team's standards document.\n\nWhich statement accurately compares Azure PowerShell and the Azure CLI?","options":["They are functionally equivalent for managing Azure resources, and teams generally choose based on the shell they already know","The Azure CLI is built for large automated deployments, while Azure PowerShell is limited to small interactive tasks","Azure PowerShell can be used only inside Azure Cloud Shell, while the Azure CLI can be installed locally","The two tools send requests through different management APIs, so their capabilities differ from tool to tool"],"answer":"A","reference":"https://learn.microsoft.com/en-us/training/modules/describe-features-tools-manage-deploy-azure-resources/2-describe-interacting-azure/","referenceTitle":"Describe tools for interacting with Azure — Microsoft Learn"}}
{"id":"RAN-AZ900-00260","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:6a4d701578ea2de94ff3f153ee761d66d4454607b715cdcb4e66ead8a7b910eb","aliases":[],"snapshot":{"statement":"A pharmaceutical distributor runs workloads in its own datacenter, in Azure, and in another public cloud. Management wants one consistent way to govern, inventory, and monitor all of these environments instead of using disconnected tools for each.\n\nWhat is the purpose of Azure Arc?","options":["To provide a browser-based shell for running interactive commands against Azure resources","To extend Azure management and governance to hybrid and multicloud resources through Azure Resource Manager","To assess on-premises virtual machines and then migrate them into Azure regions at scale","To define Azure infrastructure in declarative JSON files that make deployments repeatable and consistent"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-arc/overview","referenceTitle":"Azure Arc overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00261","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:454ab562372090246c5b88fc2d229b3275a5e97e037e668bcd2a2bd555f85a10","aliases":[],"snapshot":{"statement":"A grocery retailer will start using Azure Arc for the Windows and Linux servers in its regional datacenters. A junior engineer asks how those machines, which stay on-premises, become manageable with Azure tools.\n\nHow does Azure Arc make non-Azure resources manageable?","options":["It projects existing non-Azure resources into Azure Resource Manager so they can be managed as if they ran in Azure","It replicates each on-premises server as a complete virtual machine in the nearest Azure region","It routes datacenter network traffic through an Azure VPN gateway so the servers appear to be inside Azure","It converts each server into an ARM template that can be redeployed inside an Azure subscription"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-arc/overview","referenceTitle":"Azure Arc overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00262","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:4da855a2020b76798c21fc84eb79e0b9cb7b1d4208975c14bd806eb88d19ad4c","aliases":[],"snapshot":{"statement":"A hospital network wants to keep several database and container workloads on-premises but manage them with Azure tools. The architecture team is checking which resource types Azure Arc can manage outside Azure.\n\nWhich set of resource types can Azure Arc manage outside Azure?","options":["Azure Functions, Logic Apps, and Azure App Service instances","Azure Virtual Networks, subnets, and VPN gateways","Windows and Linux servers only; databases and Kubernetes clusters must run in Azure","Servers, Kubernetes clusters, Azure data services, and SQL Server hosted outside Azure"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/azure-arc/overview","referenceTitle":"Azure Arc overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00263","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:43530c6ec786a5fbb455c2a7c4913bdd17ee59c01012ba2fbc633b55dcff49c7","aliases":[],"snapshot":{"statement":"A mining company must keep its warehouse management system on on-premises servers for three more years, but security leadership wants those servers to follow the same policies, inventory, and monitoring as the company's Azure resources, starting now.\n\nWhich approach should the company choose?","options":["Use Azure Migrate to assess the servers and replicate them into an Azure region where the policies apply","Use ExpressRoute to create a private network connection, which will bring the connected servers under Azure Policy","Use Azure Arc to project the servers and apply the same policies, inventory, and monitoring as cloud resources","Replicate the servers' data to an Azure Storage account and apply Azure Policy to that storage account"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-arc/overview","referenceTitle":"Azure Arc overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00264","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:97f8b57ce1fc009da271ca365e68b4f1485481aa3a426a254ce08b4ed1b0dbce","aliases":[],"snapshot":{"statement":"A software vendor's engineering director wants infrastructure changes handled the way application code is: reviewed, versioned, and reproducible by anyone on the team. A junior developer asks what the practice is called.\n\nWhich statement describes infrastructure as code?","options":["Recording the manual steps needed to configure each environment in a shared operations document","Defining and deploying infrastructure through descriptive, versioned models instead of configuring environments by hand","Backing up infrastructure settings after every change so they can be restored if an environment drifts","Monitoring resource utilization and letting automation resize resources as demand changes"],"answer":"B","reference":"https://learn.microsoft.com/en-us/devops/deliver/what-is-infrastructure-as-code","referenceTitle":"What is infrastructure as code (IaC)? — Microsoft Learn"}}
{"id":"RAN-AZ900-00265","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:112f22cafaf19cb44598950631f085cbb5b1d721fa74ba117bcfa2d8fe74c1b6","aliases":[],"snapshot":{"statement":"At a food delivery platform, each test environment was configured manually by different engineers. The environments have drifted apart, deployments fail unpredictably, and nobody can reproduce them exactly. The engineering manager wants any team member to be able to rebuild an identical environment on demand.\n\nWhich practice addresses this problem?","options":["Apply resource locks to the test environments so that no one can modify their settings","Have each engineer maintain a wiki page describing the changes they made to each environment","Follow Azure Advisor recommendations to bring the environments into alignment","Adopt infrastructure as code so that every environment is generated from the same versioned definition"],"answer":"D","reference":"https://learn.microsoft.com/en-us/devops/deliver/what-is-infrastructure-as-code","referenceTitle":"What is infrastructure as code (IaC)? — Microsoft Learn"}}
{"id":"RAN-AZ900-00266","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:4aa36f644b2cf51674820db6d62424e8bf49c7659a2bb38984b85dfb4550a31b","aliases":[],"snapshot":{"statement":"A game studio redeploys its staging infrastructure frequently. The lead engineer wants the assurance that applying the same deployment definition to an environment — whether it is empty, partially configured, or already complete — always converges to the same end state.\n\nWhich IaC principle provides this assurance?","options":["Idempotence, where a given operation always produces the same result regardless of the environment's starting state","Declarative syntax, where the definition file lists the exact commands to execute, in order","Orchestration, where engineers sequence every resource creation step manually in the definition","Modularity, where every environment receives its own separate and uniquely maintained definition file"],"answer":"A","reference":"https://learn.microsoft.com/en-us/devops/deliver/what-is-infrastructure-as-code","referenceTitle":"What is infrastructure as code (IaC)? — Microsoft Learn"}}
{"id":"RAN-AZ900-00267","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:f46fe4cc37d3ffdcde1e8685306f0bfcbd90fa7f533ea1e326f4fb2d3e2fbe61","aliases":[],"snapshot":{"statement":"An e-commerce company provisions its Azure environment with a long Bash script: create the network, then the virtual machines, then the database, in a strict sequence. When a step fails mid-script, the environment is left half-built. The team wants to describe the desired end state once and let the platform determine the steps and deploy them in the correct order.\n\nWhich approach should the team adopt?","options":["Rewrite the same step-by-step commands as an Azure PowerShell script with thorough error handling","Run the existing Bash script inside Azure Cloud Shell instead of on a local terminal","Define the environment in an ARM template declaring the resources, then deploy it through Resource Manager","Perform the steps manually in the Azure portal and record them in an operations runbook"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/templates/overview","referenceTitle":"What are ARM templates? — Microsoft Learn"}}
{"id":"RAN-AZ900-00268","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:c519b112f93cd9ccd3100ad430a1f20aefad2392671b7a4c2392605a26b5c678","aliases":[],"snapshot":{"statement":"A nonprofit's IT team is studying how Azure handles resource operations behind the scenes. A volunteer with cloud experience asks for a plain definition of Azure Resource Manager.\n\nWhich statement describes Azure Resource Manager?","options":["A browser-based terminal that offers Bash and PowerShell sessions for running Azure commands","The deployment and management service for Azure — the management layer for creating, updating, and deleting resources","A service that assesses on-premises servers and applications and migrates them into Azure subscriptions","A dashboard that visualizes the health, performance, and costs of all deployed Azure resources"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/overview","referenceTitle":"What is Azure Resource Manager? — Microsoft Learn"}}
{"id":"RAN-AZ900-00269","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:f7a3a954299abaa083c62b2753e7b9ecabd4c321501a69a4dedff0ba6067a350","aliases":[],"snapshot":{"statement":"Developers at a news organization noticed that a capability available in the Azure portal is also available in the Azure CLI, Azure PowerShell, and the REST APIs. An architect asks why the tools stay consistent with each other.\n\nWhat is the reason for this consistency?","options":["All requests from Azure tools, APIs, and SDKs are handled by the same Azure Resource Manager API","The Azure portal automatically translates each of its features into scripts for the other tools","Each tool implements its own separate API, and Microsoft keeps them synchronized by policy","Azure Resource Manager processes only template deployments, so the tools share nothing beyond templates"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/overview","referenceTitle":"What is Azure Resource Manager? — Microsoft Learn"}}
{"id":"RAN-AZ900-00270","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:614ebab65c05bcb07c799fbdf3623e4b9a31d8158361849754f1159075f8e502","aliases":[],"snapshot":{"statement":"While reviewing a deployment definition at a public transit authority, an intern sees the entry \"type\": \"Microsoft.Storage/storageAccounts\" and asks what the Microsoft.Storage part is called in Azure Resource Manager terminology.\n\nWhat is Microsoft.Storage an example of?","options":["A resource group, the container for related resources","An extension resource that adds capabilities to another resource","A resource provider, a service that supplies Azure resources","A management group, which organizes Azure subscriptions"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/overview","referenceTitle":"What is Azure Resource Manager? — Microsoft Learn"}}
{"id":"RAN-AZ900-00271","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e3a2e8f6032f744ac01683f8c905a8216df9ab3be6862f82d8206ef3e930a0f6","aliases":[],"snapshot":{"statement":"A real estate firm's platform team is comparing deployment options for a new Azure environment. A new engineer asks what an ARM template actually is.\n\nWhich statement describes an ARM template?","options":["An imperative PowerShell script whose cmdlets create resources one at a time in a fixed order","A Bash script saved in Azure Cloud Shell that an engineer runs manually against each environment","A guided wizard in the Azure portal that walks an administrator through creating resources","A JSON file that declaratively defines the resources and configuration for a deployment"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/templates/overview","referenceTitle":"What are ARM templates? — Microsoft Learn"}}
{"id":"RAN-AZ900-00272","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:899647c9706d668612672314a5ef23db0e6f032d15f5eec7660d8043d66f5ab5","aliases":[],"snapshot":{"statement":"A telecom operator's deployment definition must work unchanged for development, staging, and production, which differ only in resource names, sizes, and locations. Engineers want to supply those values at deployment time rather than maintaining three separate files.\n\nWhich ARM template feature supports this?","options":["Outputs, which return values from the deployed resources after the deployment finishes running","Parameters, which supply values at deployment so the same template can be customized per environment","Resource locks, which freeze an environment's configuration after its first deployment completes","Deployment scripts, which embed PowerShell or Bash code inside the template itself"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/templates/overview","referenceTitle":"What are ARM templates? — Microsoft Learn"}}
{"id":"RAN-AZ900-00273","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:f8e6ebc1f0bd0770c8fc886d88d75b2a10b0f6eb83efef19b6ac5ae2200a12cb","aliases":[],"snapshot":{"statement":"An environmental monitoring company's Azure environment has 15 resources, several of which depend on others being created first. The team does not want to calculate the creation order or issue one command per resource.\n\nWhich ARM templates capability addresses this need?","options":["Extensibility, which lets engineers embed scripts that create each of the resources step by step","Modularity, which splits the deployment into separate templates that engineers must run in sequence","Orchestration, where Resource Manager deploys interdependent resources in the correct order and in parallel when possible","Validation, which checks that all of the resources already exist before the deployment is submitted"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/templates/overview","referenceTitle":"What are ARM templates? — Microsoft Learn"}}
{"id":"RAN-AZ900-00274","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:5b7add0f4f68b52d329896c4a2af9a642c20aca0ef1a0a17c604510461da34b0","aliases":[],"snapshot":{"statement":"A hotel chain's platform group is deciding how to write its Azure infrastructure definitions. A consultant recommended Bicep, and the CTO wants to confirm exactly what Bicep is before approving the standard.\n\nWhich statement accurately describes Bicep?","options":["A declarative language for deploying Azure resources; its files are automatically converted to ARM templates","An imperative scripting language that replaces Azure PowerShell for step-by-step deployment tasks","A separate deployment service that bypasses Azure Resource Manager to deploy resources faster","A legacy JSON format that predates ARM templates and is gradually being phased out"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/templates/overview","referenceTitle":"What are ARM templates? — Microsoft Learn"}}
{"id":"RAN-AZ900-00275","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:bb57387f4546841f134f3c00ad0fc9bdd16bc7d2ea04515076a1db8ca262fb27","aliases":[],"snapshot":{"statement":"A sports league's digital platform team has had deployments fail halfway, leaving partially created environments that engineers must clean up by hand. The team wants flawed definitions to be rejected automatically before any resource is created.\n\nWhich ARM templates capability addresses this?","options":["Tracked deployment history, which records every deployment and its parameter values in the Azure portal","Deployment scripts, which add PowerShell or Bash steps before or after resources are created","The what-if operation, which previews the changes a deployment would make before it is submitted","Built-in validation, which checks the template before deployment begins so it is less likely to stop half-finished"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/azure-resource-manager/templates/overview","referenceTitle":"What are ARM templates? — Microsoft Learn"}}
{"id":"RAN-AZ900-00276","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:e0b67647b352aa43e8b84dca0afd7a5fceed436916d36b131607d848656daa3e","aliases":[],"snapshot":{"statement":"A video streaming company runs its platform on Azure. The operations lead wants the team to follow Azure best practices without hiring external consultants: she needs a built-in tool that examines how the company's deployed resources are configured and used, and then produces personalized, actionable improvement suggestions that the team can apply or ignore. Which tool should the team use?","options":["Azure Service Health, which reports outages and maintenance events for the Azure services the company uses","Azure Cost Management, which tracks spending against budgets and forecasts future charges","Azure Advisor, which analyzes resource configuration and usage telemetry to recommend improvements","Azure Monitor, which collects metrics and logs from the company's resources and applications"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/advisor/advisor-overview","referenceTitle":"Introduction to Azure Advisor — Microsoft Learn"}}
{"id":"RAN-AZ900-00277","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:69d54d260121a4195e77ae75a401450c940a6efab78376655d698907de1a522a","aliases":[],"snapshot":{"statement":"A logistics company's cloud engineer is reviewing the Azure Advisor dashboard. One recommendation flags a configuration risk in how several production virtual machines are set up: if a component fails, the delivery-tracking applications could stop running, and Advisor proposes changes that would keep them operating through such failures. Under which Advisor recommendation category does this guidance fall?","options":["Security, which identifies threats and vulnerabilities that could lead to breaches","Cost, which finds opportunities to reduce overall Azure spending","Reliability, which helps keep applications running by flagging configuration risks","Operational excellence, which suggests improvements to process, workflow, and deployment practices"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/advisor/advisor-overview","referenceTitle":"Introduction to Azure Advisor — Microsoft Learn"}}
{"id":"RAN-AZ900-00278","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:113d3fafe9dd4335b92d096ce9f5acb1205159f785229af7515f792b457a5e65","aliases":[],"snapshot":{"statement":"The CFO of a fintech startup challenges the platform team to cut the company's Azure bill. The team already uses Microsoft Cost Management: cost analysis shows where the money goes, and budgets with alerts are in place — yet spending keeps drifting up because no one knows which resources to change. The CFO asks for concrete, reviewable actions rather than another report. Which tool should the team use?","options":["Azure Advisor, which analyzes deployed resources and recommends cost optimizations with actions you can review and apply","Microsoft Cost Management, which reports past and forecasted spending and compares it with budgets","Azure Policy, which enforces rules such as allowed SKUs on resource deployments","Azure Service Health, which reports service issues and planned maintenance events that affect the subscription"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/advisor/advisor-overview","referenceTitle":"Introduction to Azure Advisor — Microsoft Learn"}}
{"id":"RAN-AZ900-00279","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:4b2371a20319a4077e85e526595a24e76724341d6769e46f3b27cd27a0b0b2f7","aliases":[],"snapshot":{"statement":"An IT administrator at a university is triaging Azure Advisor recommendations with her team. One recommendation about a storage account is valid, but the project that will address it does not start for another two months, and she does not want the item to keep appearing as an active recommendation in the meantime. The team will revisit it once the project begins. What should she do in Advisor?","options":["Dismiss the recommendation, which marks it as not relevant to the environment","Leave it as is; Advisor eventually implements recommended changes automatically","Convert the recommendation into a Service Health alert so it stops cluttering the dashboard","Postpone the recommendation, which temporarily hides it for later review"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/advisor/advisor-overview","referenceTitle":"Introduction to Azure Advisor — Microsoft Learn"}}
{"id":"RAN-AZ900-00280","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:f9ae987eafae6d36bdfeae4e290cd4f57b721702daec78f9d78b5105e8f19dc9","aliases":[],"snapshot":{"statement":"During an architecture review at a utility company, a junior architect reads out four statements about Azure Advisor to the review board. Which statement is accurate?","options":["Advisor presents a global view of the health of all Azure services across all Azure regions","Advisor evaluates deployed Azure resources and recommends best-practice improvements in five categories","Advisor collects application telemetry in near real time and fires alerts when metric thresholds are crossed","Advisor notifies administrators about Azure service outages and upcoming planned maintenance events"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/advisor/advisor-overview","referenceTitle":"Introduction to Azure Advisor — Microsoft Learn"}}
{"id":"RAN-AZ900-00281","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:917804839f1cd01f3d70e7dcc8a41208c8c05638fc516b14d99911e256066b89","aliases":[],"snapshot":{"statement":"An online retailer runs its storefront on several Azure services. The operations manager wants a portal experience that shows active outages, upcoming planned maintenance, and health advisories — but only for the Azure services and regions the retailer actually uses, so the team is not distracted by events that cannot affect them. Which service provides this view?","options":["Azure Service Health, which provides a personalized view of the Azure services and regions you use","Azure Status, which presents a global view of the health of all Azure services across all regions","Azure Resource Health, which reports the current health of an individual resource such as one virtual machine","Azure Advisor, which recommends configuration improvements for reliability, performance, and cost"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/service-health/overview","referenceTitle":"What is Azure Service Health? — Microsoft Learn"}}
{"id":"RAN-AZ900-00282","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:0f0e08daab52be926143e9204c02f872ddf0e3a7b31fa53aca29221cf693f517","aliases":[],"snapshot":{"statement":"News outlets are reporting a widespread Azure incident. The IT director of a public-sector agency wants to check the overall condition of the Azure platform itself — every Azure service in every region, including services the agency does not use — before deciding whether to investigate the agency's own workloads. Where should the director look?","options":["Azure Service Health, which lists outages and maintenance events for the services and regions the agency uses","Azure Monitor, which collects and analyzes metrics and logs from the agency's resources and applications","Azure Resource Health, which reports the current availability of each individual resource the agency runs","Azure Status, a global view of the health of all Azure services in every region"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/service-health/overview","referenceTitle":"What is Azure Service Health? — Microsoft Learn"}}
{"id":"RAN-AZ900-00283","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:a74445d4309a206561c5ab81f2d9f686164295e4d8cc0de8861ccf007341b1b8","aliases":[],"snapshot":{"statement":"A hospital's patient portal depends on Azure services that occasionally need platform maintenance. The infrastructure team wants advance notice of upcoming maintenance events that could affect the portal's services, so that maintenance windows can be scheduled and clinical staff warned before anything happens. Which Azure Service Health area should the team monitor?","options":["Service issues, which list active incidents that are already affecting Azure services","Planned maintenance, which identifies upcoming maintenance events so you can plan for them","Health advisories, which communicate changes to Azure services that might affect your resources","Resource Health, which reports the current availability of individual resources"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/service-health/overview","referenceTitle":"What is Azure Service Health? — Microsoft Learn"}}
{"id":"RAN-AZ900-00284","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:b6f76f7633adbe34380f320d2259e896706b3031acef6d17ac0252461ccbe3a1","aliases":[],"snapshot":{"statement":"A CRM startup's services are used by customers around the clock. The on-call engineer cannot watch the Azure portal all day and wants to be notified automatically — by email and text message — whenever Azure-side events such as outages or planned maintenance affect the specific services and regions the startup depends on. What should the engineer configure?","options":["An Azure Advisor notification, which sends email when new best-practice recommendations appear for the subscription","An Azure Monitor metric alert that fires when the application servers' CPU utilization crosses a threshold","A Service Health alert, which proactively notifies you about issues and changes affecting your services","A scheduled nightly review of the Azure Status page, which lists incidents across all Azure regions"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/service-health/overview","referenceTitle":"What is Azure Service Health? — Microsoft Learn"}}
{"id":"RAN-AZ900-00285","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:e0420feacae36e43b78b85f13b51f13577bab80e680fe416450c2dfe674cb871","aliases":[],"snapshot":{"statement":"At a manufacturing plant, the monitoring application on an Azure virtual machine suddenly becomes unreachable. Before opening a support ticket, the engineer wants to check whether this specific virtual machine is unhealthy because of a problem on the Azure platform or because of something in the company's own setup. Which capability should the engineer check?","options":["Azure Resource Health, which assesses one resource and shows whether the issue is Azure's or yours","Azure Service Health, which lists service-impacting events and advisories for the services and regions the company uses","Application Insights, which collects request, failure, and dependency telemetry from the plant's web application","Azure Status, which presents the global health of all Azure services across all Azure regions"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/service-health/resource-health-overview","referenceTitle":"Azure Resource Health overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00286","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:33bb93ee8c55b322a1c1e35173ab30789642a6edd78d62eb7b7dd23d8302e60d","aliases":[],"snapshot":{"statement":"A cloud architect at an international NGO is onboarding new administrators. As part of the training material, she must document the three components of Azure Service Health and the scope that each one covers. Which mapping is correct?","options":["Azure Status: health of the services you use; Service Health: global health of all Azure services; Resource Health: health of individual resources","Azure Status: global health of all Azure services; Service Health: health of individual resources; Resource Health: health of the services and regions you use","Azure Status: global health of all Azure services; Service Health: health of the services and regions you use; Resource Health: health of individual resources","Azure Status: health of individual resources; Service Health: global health of all Azure services; Resource Health: health of the services and regions you use"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/service-health/overview","referenceTitle":"What is Azure Service Health? — Microsoft Learn"}}
{"id":"RAN-AZ900-00287","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:5b5ae286ff41cecab1938998155ba4ca8e6c9b8cf04c7593e0d5bf38a5e8b42c","aliases":[],"snapshot":{"statement":"An insurance company is defining its monitoring strategy. The platform team wants a single service that gathers both metrics and logs from the company's Azure resources and applications, stores that data centrally, and lets engineers analyze it through dashboards and queries and respond automatically with alert rules. Which service should the team use?","options":["Azure Advisor, which scans your deployed resources and recommends best-practice improvements","Azure Service Health, which communicates Azure service outages and planned maintenance events to you","Azure Policy, which audits and enforces organizational rules on new resource deployments","Azure Monitor, which collects, analyzes, and acts on telemetry from resources and applications"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/fundamentals/overview","referenceTitle":"Azure Monitor overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00288","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:05c1fd0126db5f39b7a02eabb430925cf8d3223e1f3bd23bce07a159a37309cd","aliases":[],"snapshot":{"statement":"During a monitoring design session at a game studio, two engineers are documenting how Azure Monitor organizes the data it collects. One of them asks for the precise difference between the two data platforms. Which description is correct?","options":["Metrics are numerical values sampled at regular intervals and stored in a time-series database; logs are timestamped event records containing structured or free-form text","Metrics are timestamped event records that can contain structured or free-form text; logs are numerical values sampled at regular intervals in a time series","Both metrics and logs are stored as numerical time series, and the difference between the two platforms is only in how charts display them","Metrics are kept in a Log Analytics workspace, while logs are kept alongside their applications in Application Insights components"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/fundamentals/data-platform","referenceTitle":"Azure Monitor data platform — Microsoft Learn"}}
{"id":"RAN-AZ900-00289","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:5ae67a190ea9bb836667160864565b51aea4874a12b1a49614b8deb7dbfd6a53","aliases":[],"snapshot":{"statement":"A security analyst at an e-commerce company needs to explore the log data that Azure Monitor has collected from the company's Azure resources — starting with a simple filter for events in the last hour, then moving on to trend analysis over the past week. Which tool in the Azure portal should the analyst use to write and run these queries?","options":["Log Analytics, the tool for writing and running queries against Azure Monitor Logs","Application Insights, which monitors the performance and usage of the company's web applications","Azure Monitor alerts, which send notifications when conditions that you define are met","Azure Advisor, which lists best-practice recommendations for the company's deployed resources"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/logs/log-analytics-overview","referenceTitle":"Overview of Log Analytics in Azure Monitor — Microsoft Learn"}}
{"id":"RAN-AZ900-00290","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:559f4846347e04b7e56f6f48eb21c8084048aeab8e20a8c32bb228728dec17fe","aliases":[],"snapshot":{"statement":"A junior engineer at an energy company has just been granted access to the Log Analytics tool in the Azure portal. Before writing her first log query, she asks a colleague which query language is used there. What is the correct answer?","options":["Transact-SQL (T-SQL), the query language used by Azure SQL databases","Python scripts, which the Log Analytics engine executes against the workspace","Kusto Query Language (KQL), the query language used in Log Analytics","Azure CLI commands, which manage Azure resources from the command line"],"answer":"C","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/logs/log-analytics-overview","referenceTitle":"Overview of Log Analytics in Azure Monitor — Microsoft Learn"}}
{"id":"RAN-AZ900-00291","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:56bf1205f32796f42e3f7e0a11f90f218ae0cb80c585dcf5412ea054797c5dcb","aliases":[],"snapshot":{"statement":"Investigators at a bank are reconstructing a security incident. They need a single experience where they can write queries that combine log data collected from virtual machines, storage accounts, and the bank's web application telemetry — all stored by Azure Monitor. Which capability should the investigators use?","options":["Application Insights, which focuses on the performance and usage of the bank's web application","Azure Monitor Metrics, which charts numerical values sampled at regular intervals over time","Azure Advisor, which provides security recommendations for the resources the bank has deployed","Log Analytics, where queries run across the logs Azure Monitor collects from multiple sources"],"answer":"D","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/logs/log-analytics-overview","referenceTitle":"Overview of Log Analytics in Azure Monitor — Microsoft Learn"}}
{"id":"RAN-AZ900-00292","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:3c93d96e84068cb248b707b24d5a8b036760b4b53e89cfbe9fbabe586497bc1e","aliases":[],"snapshot":{"statement":"A retail chain has grown by acquisition, and each business unit has its own Azure subscription. Analysts currently repeat every query in each subscription's portal view because the log data lands in separate places. Central IT wants all log data from across the organization's subscriptions collected into one shared store that a single query can target. What should Central IT set up?","options":["A separate Log Analytics workspace in every subscription, to keep each environment's data isolated","A single Log Analytics workspace that collects log data from resources across the subscriptions","One Application Insights component per application, gathering each app's telemetry separately","A Service Health dashboard that aggregates service events for all subscriptions"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/logs/log-analytics-workspace-overview","referenceTitle":"Log Analytics workspace overview — Microsoft Learn"}}
{"id":"RAN-AZ900-00293","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:f039774e6c41e8d0c89b0c3309a337682ae073e8741165bca32c793e181624e1","aliases":[],"snapshot":{"statement":"A travel booking platform is designing its monitoring pipeline. The lead engineer sketches the flow: telemetry is gathered from applications, operating systems, and the network layer; the data is stored centrally; engineers analyze it by writing queries; and alert rules evaluate it and trigger notifications when conditions are met. Which combination of Azure Monitor components implements this flow?","options":["Azure Advisor collects the telemetry, the Azure Status page stores it, and Service Health queries it using the KQL query language","Azure Monitor gathers metrics and logs, Log Analytics queries them, and alert rules with action groups fire when conditions are met","Application Insights collects all resource telemetry, Log Analytics queries only metrics, and Azure Policy triggers the alert actions","Log Analytics collects the telemetry, Azure Monitor alerts store it as metrics, and Azure Status displays all the dashboards"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-monitoring-tools-azure/4-describe-azure-monitor/","referenceTitle":"Describe Azure Monitor — Microsoft Learn"}}
{"id":"RAN-AZ900-00294","exam":"AZ900","topics":["management"],"difficulty":"easy","lang":"en","version":1,"checksum":"sha256:9678bf20433cd7df7388b852350384a91176f6ebf4efe65b1b04478d221234bd","aliases":[],"snapshot":{"statement":"A marketing analytics firm runs a virtual machine that processes nightly data loads. Operations wants an automatic email to the team's mailbox whenever the VM's CPU utilization stays above 80 percent for five minutes, so engineers can intervene before the load window is missed. Which capability should they configure?","options":["An Azure Monitor metric alert that watches the VM's CPU and emails the team when it fires","A Service Health alert that reports Azure service outages and other platform incidents affecting the VM's region","An Azure Advisor cost recommendation that proposes resizing the virtual machine to a smaller size","An Application Insights availability test that sends periodic web requests to the workload's public endpoint"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/alerts/alerts-overview","referenceTitle":"Overview of Azure Monitor alerts — Microsoft Learn"}}
{"id":"RAN-AZ900-00295","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:d304eb7f5c69e6e205866653b6f79894975213333931cc5ac4f6bbfe83068f29","aliases":[],"snapshot":{"statement":"A media company operates dozens of alert rules across its Azure environment. Whenever any alert fires, the same response must happen: two on-call engineers receive email and text messages, and a Logic App opens a support ticket. The team wants to define this response once and attach it to current and future alert rules. What should they create?","options":["A separate alert rule for each engineer, each rule with its own notification settings","An Azure Monitor workbook that displays fired alerts for the team to review every morning","A diagnostic setting that streams fired alert records to a storage account for later distribution","An action group that defines the notifications and actions once for all alert rules"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-monitoring-tools-azure/4-describe-azure-monitor/","referenceTitle":"Describe Azure Monitor — Microsoft Learn"}}
{"id":"RAN-AZ900-00296","exam":"AZ900","topics":["management"],"difficulty":"hard","lang":"en","version":1,"checksum":"sha256:07263737785db963c29a0d7ac2ce21e60ba248ee1075c318bcb3957935d22460","aliases":[],"snapshot":{"statement":"A video platform's reliability team has learned that a specific database connection error message appearing across the logs of several backend services is an early sign of an outage. They want an alert that fires whenever that text pattern shows up in the log data collected in their Log Analytics workspace. Which type of alert should they create?","options":["A metric alert on each backend service, watching a numerical value that changes in near real time","A log alert that queries the workspace and fires when the error pattern is found","A Service Health alert that tracks Azure platform incidents and maintenance in the services' regions","An Application Insights availability test that checks whether the public website endpoint responds"],"answer":"B","reference":"https://learn.microsoft.com/en-us/training/modules/describe-monitoring-tools-azure/4-describe-azure-monitor/","referenceTitle":"Describe Azure Monitor — Microsoft Learn"}}
{"id":"RAN-AZ900-00297","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:ea22a18a29ef657088c1c372cfcf98441f04c01551689a1334871027955bdc25","aliases":[],"snapshot":{"statement":"An airline's booking website becomes slow during fare promotions. The web team wants to understand the application from the users' perspective: request rates, response times, failure rates, page load times, and how usage trends across user sessions. Which capability should the team enable for the website?","options":["Log Analytics, for writing ad hoc queries across all log data collected in the workspace","Azure Service Health, which reports Azure-side outages and maintenance for the services in use","Application Insights, which monitors application performance and usage such as requests, failures, and page loads","Azure Advisor performance recommendations, which suggest resource changes that can speed up applications"],"answer":"C","reference":"https://learn.microsoft.com/en-us/training/modules/describe-monitoring-tools-azure/4-describe-azure-monitor/","referenceTitle":"Describe Azure Monitor — Microsoft Learn"}}
{"id":"RAN-AZ900-00298","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:85b9c37f17cae2fc406e0b0a84578f5ca6cfcd09842c13c1ff04cd82aa48aa71","aliases":[],"snapshot":{"statement":"A government services portal must stay reachable around the clock, but almost nobody uses it at night. The operations team wants to know promptly if the public sign-in endpoint stops responding — including during low-traffic hours, when there are too few real users for their activity to reveal the problem. Which Application Insights capability should they use?","options":["Availability tests, which send synthetic requests at regular intervals and alert if the endpoint stops responding","A log alert that counts failed requests found in the portal's Log Analytics workspace","A Service Health alert that reports platform events for the region where the portal is hosted","An Azure Advisor reliability recommendation that flags configuration risks in the portal's resources"],"answer":"A","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/app/availability","referenceTitle":"Application Insights availability tests — Microsoft Learn"}}
{"id":"RAN-AZ900-00299","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:391818f8464f6f0139ca5514463469724d8a15879d914850395de796f26d59dd","aliases":[],"snapshot":{"statement":"A healthcare provider runs its appointment application on Azure App Service. The development team wants Application Insights telemetry — requests, dependencies, and failures — but a release freeze means no application code can change right now. How can the team start collecting this telemetry?","options":["Install the Azure Monitor agent on the underlying hosts; it automatically captures request and dependency telemetry from the application","Wait until the freeze ends and then add the Application Insights SDK to the application code","Create a Service Health alert for the application and review the event notifications","Enable the Application Insights agent for the supported runtime, which collects telemetry without code changes"],"answer":"D","reference":"https://learn.microsoft.com/en-us/training/modules/describe-monitoring-tools-azure/4-describe-azure-monitor/","referenceTitle":"Describe Azure Monitor — Microsoft Learn"}}
{"id":"RAN-AZ900-00300","exam":"AZ900","topics":["management"],"difficulty":"medium","lang":"en","version":1,"checksum":"sha256:df50ca0650b4cb0cd80fdee5b1faa1bf35aff359643c02bcf8e3010e560e4bc0","aliases":[],"snapshot":{"statement":"The licensing manager at a manufacturing company is reviewing monitoring expenses. The platform team already uses Azure Monitor to collect metrics and logs from the company's infrastructure. The manager asks whether Application Insights — which the developers want for the company's web applications — is yet another monitoring product that must be purchased and managed separately. What is the correct answer?","options":["Yes — it is a separate observability product, because Azure Monitor covers only infrastructure metrics","No — Application Insights is an application performance monitoring (APM) feature of Azure Monitor","No — it is included with Azure Service Health, which tracks whether applications are available","Yes — it is a Microsoft Defender for Cloud plan focused on application security"],"answer":"B","reference":"https://learn.microsoft.com/en-us/azure/azure-monitor/app/app-insights-overview","referenceTitle":"Application Insights OpenTelemetry observability overview — Microsoft Learn"}}