What is end_date in Airflow: Definition and Usage

In Airflow, end_date is a parameter that defines the last date and time a task or DAG should run. It acts like a stop sign, telling Airflow not to schedule any runs after this date.

How It Works

Think of end_date in Airflow as a calendar deadline for your tasks or workflows. When you set an end_date, Airflow will schedule runs only up to that date and no further. This helps control how long your workflows keep running.

Imagine you have a daily report that should only run until the end of the year. By setting end_date to December 31st, Airflow will automatically stop scheduling the report after that day, so you don’t have to manually turn it off.

Example

This example shows a DAG with a start_date and an end_date. The DAG runs daily but stops after the end_date.

python

from airflow import DAG
from airflow.operators.dummy import DummyOperator
from datetime import datetime

with DAG(
    dag_id='example_end_date_dag',
    start_date=datetime(2024, 1, 1),
    end_date=datetime(2024, 1, 5),
    schedule_interval='@daily',
    catchup=True
) as dag:
    task = DummyOperator(task_id='dummy_task')

Output

The DAG will run daily from 2024-01-01 to 2024-01-05 and then stop scheduling further runs.

When to Use

Use end_date when you want to limit how long a DAG or task runs. This is useful for temporary workflows, seasonal jobs, or data pipelines that only need to run for a fixed period.

For example, if you have a data cleanup job that should only run during a migration window, setting an end_date ensures it stops automatically after the migration ends.

Key Points

• end_date stops scheduling after the specified date and time.
• It works together with start_date and schedule_interval to control DAG runs.
• If end_date is not set, the DAG or task can run indefinitely.
• Setting end_date helps avoid unwanted runs and resource use.

Key Takeaways

end_date defines the last date a DAG or task will run in Airflow.

It prevents scheduling runs after the specified date, acting like a stop sign.

Use it to control temporary or time-limited workflows easily.

Without end_date, workflows can run indefinitely if scheduled.

It works with start_date and schedule_interval to manage run timing.