If you want to create a nice dashboard that displays the statuses of the most recent Airflow DAG runs, you will need to retrieve them from the Airflow database:
Table of Contents
SELECT
dag_run.dag_id,
dag_run.state
FROM dag_run
INNER JOIN (SELECT
dag_id,
MAX(execution_date) AS date
FROM dag_run
GROUP BY dag_id) mx
ON
dag_run.dag_id = mx.dag_id
AND dag_run.execution_date = mx.date
JOIN dag ON dag.dag_id = dag_run.dag_id AND is_active = 1 AND is_paused = 0
This query retrieves the most recent DAG run and returns its id and state. Additionally, it filters out the paused DAGs, so we don’t pollute the results with things that don’t run in production right now.
Of course, creating a dashboard of Airflow DAG runs takes some time, so you may prefer to use the project we have open-sourced: https://github.com/Wikia/discreETLy. In addition to DAG execution times and statuses, it also displays Athena tables’ last update time and their descriptions.
Get Weekly AI Implementation Insights
Join engineering leaders who receive my analysis of common AI production failures and how to prevent them. No fluff, just actionable techniques.