DAG

A DAG (Directed Acyclic Graph) is a collection of tasks organized so that each task has clear dependencies, and the graph never loops back on itself (“acyclic” — no cycles).

In Airflow specifically, a DAG defines:

• What tasks exist (e.g. create topic, produce events, run Spark job, verify results)
• What order they run in (dependencies, expressed like task_a >> task_b)
• When the whole thing runs (a schedule, or manual trigger)

Example dependency chain:

t1_create_topic >> t2_produce >> t3_spark >> t4_verify

This means t2 only starts after t1 succeeds, and so on down the chain.

See also: Airflow