A data engineer builds a Lakeflow Job with four tasks: ingest_bronze, clean_silver, aggregate_gold, and refresh_dashboard. clean_silver must start only after ingest_bronze finishes, and refresh_dashboard only after aggregate_gold completes, while independent branches should run in parallel. How should the job be structured?
Answer & explanation
Correct answer: B
WHY B: A Lakeflow Job is a DAG of tasks; declaring depends_on on each downstream task makes the job run tasks in dependency order and parallelize independent branches automatically. WHY NOT A: Jobs do express inter-task dependencies via depends_on; collapsing everything into one notebook discards orchestration and is unnecessary. WHY NOT C: Ordering comes from task dependencies, not from staggering per-task schedule times, which is fragile and not how a job DAG works. WHY NOT D: The continuous trigger controls when the job starts (always-on), not how tasks are ordered; it does not infer a DAG from inputs. WHY NOT E: A for-each task fans a single task over a collection of inputs; it does not sequence four distinct heterogeneous tasks.