Lesson 23 · Senior Airflow judgment
Mock senior Airflow interview pack
A retrieval-first mock pack for turning the whole Airflow course into clear, interview-grade answers you can say under pressure.
Your win: run a realistic senior Airflow mock interview on yourself, score your answers honestly, and turn weak spots into a focused review plan.
In plain English Plain English: answer out loud before checking the answer points. Recognition is not recall.
What this mock is really training
This page is not only testing whether you remember facts. It is training whether you can speak clearly under pressure about orchestration boundaries, time semantics, generated DAG contracts, reliability trade-offs, and architecture judgment.
That is why this mock works best if you treat it like a real interview. Answer first. Then check the points. Then answer again more cleanly. The second answer is where a lot of learning happens.
- Round 1 · Airflow truthfulness and boundaries — 6 questions
- Round 2 · Scheduling, generation, and contract reasoning — 6 questions
- Round 3 · Reliability, architecture, and senior review judgment — 6 questions
- Time target — 60–90 seconds per answer
The answer checklist A strong answer should usually include: (1) the mechanism, (2) the main trade-off or limit, and (3) a repo-grounded consequence or example.
Round 1 · Airflow truthfulness and boundaries
- What does Airflow really own in this repo?
- Why is “Airflow runs the data platform” too vague?
- Why is it useful to separate orchestration from Spark execution when debugging?
- How would you summarize the repo’s Airflow role in one honest paragraph?
- Why should strong answers mention CDC, Spark, Kubernetes, and the warehouse?
- How would you explain the difference between control-plane ownership and execution ownership?
Round 1 · Strong answer points
- Boundary clarity. Airflow owns orchestration, scheduling, state, and alerting.
- Truthfulness. That does not mean it owns compute, storage, or upstream CDC plumbing.
- Debugging value. Separating layers helps you fix the right thing faster.
- Concise summary. Airflow coordinates the downstream workflow layer of a larger data path.
- Adjacency. Neighboring systems still own processing, storage, and substrate behavior.
- Ownership model. Control-plane and execution-plane concerns are connected but not identical.
Round 2 · Scheduling, generation, and contract reasoning
- Why are Airflow scheduling semantics a correctness topic, not just a cron topic?
- How would you explain logical date vs run time to a teammate?
- Why is `catchup=False` a posture decision here?
- What is the real value of generated DAGs in this repo?
- What is the hidden cost of generated DAGs?
- Why is the drift check more important than it first sounds?
Round 2 · Strong answer points
- Time semantics. Interval ownership affects correctness, replay, and alert interpretation.
- Mental model. Logical date marks interval start; runs fire after interval end.
- Operational posture. `catchup=False` avoids accidental historical floods by default.
- Generation win. Rules and templates centralize repeated patterns and enforcement.
- Generation cost. Indirection increases tracing and review complexity.
- Drift safety. CI ensures committed DAGs truly match rules plus templates.
Round 3 · Reliability, architecture, and senior review judgment
- Why can a green DAG still mean a failed pipeline?
- What does this repo do to protect correctness, not just liveness?
- Why do executor and Spark boundaries matter when debugging?
- How should you describe the triggerer honestly in this repo?
- What would you ask first in a risky Airflow PR?
- What separates a senior Airflow answer from a doc-only answer?
Round 3 · Strong answer points
- False-green risk. Orchestration success does not guarantee data correctness.
- Correctness controls. Validation tasks and Great Expectations protect trusted output.
- Boundary debugging. Spark/Kubernetes failures are not automatically Airflow logic failures.
- Triggerer honesty. Present in Airflow 3 architecture, but real DAG usage here has no sensors.
- Review lens. Start with ownership boundary, time semantics, contract surface, and correctness impact.
- Judgment layer. Senior answers add truthfulness, trade-offs, and operational consequence.
Self-scoring rubric
| Score | Meaning |
|---|---|
| 0 | I could not explain it without notes. |
| 1 | I gave fragments, but the answer was incomplete or fuzzy. |
| 2 | I explained the main idea, but missed the trade-off, limit, or repo consequence. |
| 3 | I gave a strong, clear answer with mechanism + trade-off/limit + repo consequence. |
Recovery map — where to go next Missed fundamentals? Revisit Lessons 1–4. Missed scheduling semantics? Revisit Lessons 5–8 plus 18. Missed repo-grounded architecture? Revisit Lessons 9–12 plus 17, 19, and 21. Missed reliability and judgment? Revisit Lessons 13–16 plus 20–22.
Q1. The main rule of this mock pack is…
Retrieval first is what turns recognition into interview recall.
What are the three ingredients of a strong senior Airflow answer?
recall, then click to reveal
A clear mechanism, the important trade-off or limit, and a repo-grounded consequence or example.
If you want, I can now run this as a live mock senior Airflow interview and grade you answer by answer. Ask me.
Sources. This pack synthesizes the whole course into retrieval practice.