A streaming media company is designing its ingestion layer on the Databricks Data Intelligence Platform. It has three sources: - A legacy billing system that delivers a complete extract of all open invoices every Sunday night - A partner that drops small CSV files into an Amazon S3 bucket several times per hour - A website that emits click events continuously to a message bus Which mapping of ingestion pattern to source is correct?

1. A. Billing extract: streaming ingestion; partner CSVs: reload the entire bucket on every run; click events: weekly scheduled batch 💬 A
2. B. All three sources: nightly batch with full reprocessing, because batch processing keeps track of what has already been processed 💬 B
3. C. Billing extract: batch load; partner CSVs: incremental loading of only the newly arrived files; click events: streaming ingestion ✓ C
4. D. Billing extract: incremental load keyed by invoice ID; partner CSVs: stateful streaming deduplication; click events: hourly batch of the previous hour 💬 D

Reveal answer

Correct answer: Billing extract: batch load; partner CSVs: incremental loading of only the newly arrived files; click events: streaming ingestion

Reference: https://docs.databricks.com/aws/en/data-engineering/batch-vs-streaming