seanghaeli opened a new issue, #77060: URL: https://github.com/apache/airflow/issues/77060
Today the only way to stop an asset watcher is to pause every Dag scheduled on its asset or remove it from the Dag. Both delete its trigger (`add_asset_trigger_references` in `airflow-core/src/airflow/dag_processing/collection.py`), so its failure history goes too (#77054), and pausing the Dags also stops any other assets or schedules they use. When a watcher keeps failing because the system it watches is down, it would help to pause just that watcher until it's fixed, and resume it later with its history intact. Suggestion: a pause flag on the asset watcher, set through the API and UI. The triggerer doesn't run a paused watcher's trigger, and cleanup doesn't delete it. Watchers on one shared trigger need a rule: the trigger keeps running while any of its watchers isn't paused. With #76567, the Triggers page is a natural place for the pause button. Done when: a watcher can be paused and resumed through the API (and UI), a paused watcher's trigger isn't run or deleted, and its failure history is still there after it resumes, with tests. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
