Reviewed by Aditya Kumar · Last reviewed 2026-03-24
**Architectural Logic**: @daily vs once represent fundamentally different scheduling paradigms with distinct cost and operational implications. **@daily (Schedule Interval)**: DAG runs at fixed cron intervals—each execution processes a discrete schedule interval (e.g.,...
This medium-level SQL question appears frequently in data engineering interviews at companies like Google. While less common, it tests deeper understanding that distinguishes strong candidates. Mastering the underlying concepts (airflow) will help you answer variations of this question confidently.
Break this problem into components. Identify the core trade-offs involved, then walk the interviewer through your reasoning step by step. Demonstrate awareness of edge cases and production considerations - this is what separates good answers from great ones.
Architectural Logic: @daily vs once represent fundamentally different scheduling paradigms with distinct cost and operational implications. @daily (Schedule Interval): DAG runs at fixed cron intervals—each execution processes a discrete schedule interval (e.g., midnight-to-midnight). Produces deterministic runs; catchup can trigger historical backfills (cost explosion if catchup=True on long-running DAGs). Once: Single execution—manual trigger or one-off; no recurrence. Zero baseline scheduling cost. Why It Matters: @daily is for recurring batch pipelines (daily aggregations, data refreshes); once is for ad-hoc, backfills, or DAGs that shouldn't run on schedule. Scalability: @daily at scale requires careful catchup control; thousands of DAGs with @daily increase scheduler load. Cost: Unintended catchup on @daily can 10x compute costs; always set catchup=False unless backfill is explicit. Use idempotent tasks so reruns are safe.
Red Flag: Leaving catchup=True on @daily DAGs without understanding backfill semantics—can trigger thousands of historical runs overnight. Pro-Move: Use catchup=False + explicit backfill DAGs with date parameters for controlled historical processing.
Some links below are affiliate links. If you buy through them we may earn a small commission at no extra cost to you — it helps keep DataEngPrep free.
According to DataEngPrep.tech, this is one of the most frequently asked SQL interview questions, reported at 1 company. DataEngPrep.tech maintains an editor-reviewed database of 1,863 data engineering interview questions across 7 categories.