Daniel Standish
Daniel Standish
### Body We need to determine what to do about this warning: ``` /Users/dstandish/.virtualenvs/2.4.0/lib/python3.8/site-packages/flask_sqlalchemy/__init__.py:872 FSADeprecationWarning: SQLALCHEMY_TRACK_MODIFICATIONS adds significant overhead and will be disabled by default in the future. Set it...
They do get picked up like once or twice but pretty quickly somehow things get into a state where the scheduler thinks there is a task runninging in the dag...
Previously we relied on SerializedBaseOperator.serialize_operator for serialization of all Operator objects but this is no bueno because it does not work for mapped operator. Instead we use BaseSerialization.serialize, which calls...
This is needed because we do not ser the dag on Operator objects.
This is necessary because Operator does not serialize quite right. It doesn't (either completely or at all) include the dag object or its groups. So we can't properly determine 'needs...
Alright so this is the mothership PR where I have the whole shebang working, and then from it I cherry pick the bits that can be merged independently.
Implement core functionality in _run_raw_task for AIP-44 / database isolation. Followed by other PRs from https://github.com/apache/airflow/pull/37851.
Previously, there was a lot of bad stuff happening around try_number. We incremented it when task started running. And because of that, we had this logic to return "_try_number +...
If we don't set a timeout, it may hang indefinitely if there's a network issue.