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?

1. A. 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 ✓ A
2. B. 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 💬 B
3. C. A cron schedule firing every five minutes that runs COPY INTO over the whole directory, accepting that most runs find nothing new 💬 C
4. D. A scheduled job whose Trigger.AvailableNow step runs alongside a parallel always-on stream, because AvailableNow requires an active query to work 💬 D

Reveal answer

Correct answer: 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

Reference: https://docs.databricks.com/aws/en/jobs/file-arrival-triggers