Reputation: 1581
I know Airflow distinguishes between manual and scheduled triggers of a DAG - the pattern of their ID is different, and also the UI in Tree View shows an outlined circle for one but not the other.
I have a DAG that uses a Python callable as on_failure_callback
to emit failure alert. I now want to modify this DAG so that it will emit one type of alert when a manually triggered run fails, while emitting a different alert when the run was triggered by the scheduler.
I can do this by simply parsing the string in {{ execution_date }}
. However, that seems hacky. Is there a flag that I can check instead?
Upvotes: 0
Views: 1313
Reputation: 74
Seems very related to this question: Can I programmatically determine if an Airflow DAG was scheduled or manually triggered?
Looks like you can look up the run_id and parse that instead which might be slightly better.
Upvotes: 1