54 real Orchestration questions from the Data Engineering bank, as asked in Indian campus drives and tech interviews. Every question has a verified answer and an AI-tutor explanation on placd — free to start.
1. What is Airflow DAG?
Junior
A.A Dagster abstraction where you declare the data assets a pipeline produces and their dependencies, and Dagster derives the execution graph.
B.A per-task rule (all_success, one_failed, all_done) controlling when a task fires based on the states of its upstream tasks.
C.The Prefect component that executes tasks, configurable for sequential, concurrent, or distributed (Dask, Ray) execution.
D.a directed acyclic graph of tasks defining execution order
A.Airflow DAG — A Dagster abstraction where you declare the data assets a pipeline produces and their dependencies, and Dagster derives the execution graph.
B.Airflow DAG — A Dagster component that polls external state and launches runs when conditions are met, the event-driven complement to schedules.
C.Airflow DAG — A single execution instance of a DAG for a given logical date or trigger, containing the task instances and their states.
D.Airflow DAG — a directed acyclic graph of tasks defining execution order
A.Airflow operator — A Dagster abstraction where you declare the data assets a pipeline produces and their dependencies, and Dagster derives the execution graph.
B.Airflow operator — A per-task rule (all_success, one_failed, all_done) controlling when a task fires based on the states of its upstream tasks.
C.Airflow operator — a template defining one task's work (PythonOperator, BashOperator, sensor)
D.Airflow operator — Designing pipelines around the data assets produced and consumed rather than around task ordering, the model behind Dagster assets and Airflow assets.
A.sensor — A Dagster abstraction where you declare the data assets a pipeline produces and their dependencies, and Dagster derives the execution graph.
B.sensor — a directed acyclic graph of tasks defining execution order
C.sensor — an operator that waits for an external condition before succeeding
D.sensor — A task designed so that rerunning it with the same inputs produces the same result without duplicating data, essential for safe retries and backfills.
A.catchup=False — Two sensor modes: poke holds a worker slot while checking, reschedule frees the slot between checks; deferrable triggers improve on both.
B.catchup=False — skipping the run of missed historical intervals when a DAG is first enabled
C.catchup=False — A per-task rule (all_success, one_failed, all_done) controlling when a task fires based on the states of its upstream tasks.
D.catchup=False — An Airflow construct that visually and logically groups related tasks inside a DAG, replacing SubDAGs without their scheduling drawbacks.
A.XCom — Designing pipelines around the data assets produced and consumed rather than around task ordering, the model behind Dagster assets and Airflow assets.
B.XCom — Airflow's mechanism for passing small amounts of data between tasks
C.XCom — The Prefect component that executes tasks, configurable for sequential, concurrent, or distributed (Dask, Ray) execution.
D.XCom — A per-task rule (all_success, one_failed, all_done) controlling when a task fires based on the states of its upstream tasks.
A.Airflow scheduler — A task designed so that rerunning it with the same inputs produces the same result without duplicating data, essential for safe retries and backfills.
B.Airflow scheduler — the component that scans DAGs and queues eligible tasks
C.Airflow scheduler — Airflow's mechanism for passing small amounts of data between tasks
D.Airflow scheduler — a template defining one task's work (PythonOperator, BashOperator, sensor)
A.Airflow's mechanism for passing small amounts of data between tasks
B.A task designed so that rerunning it with the same inputs produces the same result without duplicating data, essential for safe retries and backfills.
C.A resilience pattern that retries a failed task after increasing delays, reducing load on a struggling dependency while improving success odds.
D.An Airflow construct that visually and logically groups related tasks inside a DAG, replacing SubDAGs without their scheduling drawbacks.
20. Which term means: "An Airflow construct that visually and logically groups related tasks inside a DAG, replacing SubDAGs without their scheduling drawbacks."?
A.TaskGroup — A single execution instance of a DAG for a given logical date or trigger, containing the task instances and their states.
B.TaskGroup — An Airflow construct that visually and logically groups related tasks inside a DAG, replacing SubDAGs without their scheduling drawbacks.
C.TaskGroup — A task designed so that rerunning it with the same inputs produces the same result without duplicating data, essential for safe retries and backfills.
D.TaskGroup — the component that scans DAGs and queues eligible tasks
A.A Dagster abstraction where you declare the data assets a pipeline produces and their dependencies, and Dagster derives the execution graph.
B.A Dagster component that polls external state and launches runs when conditions are met, the event-driven complement to schedules.
C.An Airflow concurrency-limiting mechanism that caps how many task instances tagged to it can run at once, protecting shared resources.
D.Designing pipelines around the data assets produced and consumed rather than around task ordering, the model behind Dagster assets and Airflow assets.
23. Which term means: "An Airflow concurrency-limiting mechanism that caps how many task instances tagged to it can run at once, protecting shared resources."?
A.pool — An Airflow construct that visually and logically groups related tasks inside a DAG, replacing SubDAGs without their scheduling drawbacks.
B.pool — A task designed so that rerunning it with the same inputs produces the same result without duplicating data, essential for safe retries and backfills.
C.pool — A Prefect-decorated Python function representing a workflow of tasks, with dynamic, native-Python control flow and observability.
D.pool — An Airflow concurrency-limiting mechanism that caps how many task instances tagged to it can run at once, protecting shared resources.
26. Which term means: "A Dagster abstraction where you declare the data assets a pipeline produces and their dependencies, and Dagster derives the execution graph."?
A.software-defined asset — An Airflow construct that visually and logically groups related tasks inside a DAG, replacing SubDAGs without their scheduling drawbacks.
B.software-defined asset — A Dagster abstraction where you declare the data assets a pipeline produces and their dependencies, and Dagster derives the execution graph.
C.software-defined asset — A task designed so that rerunning it with the same inputs produces the same result without duplicating data, essential for safe retries and backfills.
D.software-defined asset — A Dagster component that polls external state and launches runs when conditions are met, the event-driven complement to schedules.
29. Which term means: "A Prefect-decorated Python function representing a workflow of tasks, with dynamic, native-Python control flow and observability."?
A.Prefect flow — A resilience pattern that retries a failed task after increasing delays, reducing load on a struggling dependency while improving success odds.
B.Prefect flow — A Prefect-decorated Python function representing a workflow of tasks, with dynamic, native-Python control flow and observability.
C.Prefect flow — An Airflow concurrency-limiting mechanism that caps how many task instances tagged to it can run at once, protecting shared resources.
D.Prefect flow — Two sensor modes: poke holds a worker slot while checking, reschedule frees the slot between checks; deferrable triggers improve on both.
Showing 30 of 54 Orchestration questions — the full set, with answers, explanations and an AI tutor on every question, is inside.
Free to start
Answers, AI explanations, and a scored voice mock interview
Sign up free to check your answers with explanations, ask the AI tutor anything on any question, and take one full AI mock interview — scored like a real panel.