gmsantos commented on PR #23711: URL: https://github.com/apache/airflow/pull/23711#issuecomment-1165007592
@jedcunningham I got some progress here. The change that is failing the tests happens in this [change](https://github.com/apache/airflow/pull/23711/commits/2eaca8f2e390dd6e3c8874b4847d416f83c07d9b ) to include the dag_mount. I think in some cases we do not need the dag mount in the scheduler: - [1] if the DAGs are loaded from examples or baked with the Airflow image (no matter what, the DAGs will be available to the scheduler) - [2] when the DAG processor is enabled and we are not using a local executor, neither use DAGs persistence or gitSync (the mount is moved to dag processor deployment then) - [2] when the DAG processor is enabled and we are using a local executor, but we neither use DAGs persistence or gitSync (it falls on the case [1], DAGs are already here) We need to mount the DAGs to the scheduler when: - [3] DAG processor is disabled, and we use either DAGs persistence or gitSync - [4] if we use a local executor, we need to have the DAGs mount if we use either DAGs persistence or gitSync, no matter if DAG processor is disabled or enabled -- 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]
