potiuk commented on pull request #13832: URL: https://github.com/apache/airflow/pull/13832#issuecomment-765603917
> > How would that solve it ? From what I understand it would simply result in a different exception thrown - but the flow will remain the same? > > @potiuk If you check [taskinstance.py](https://github.com/apache/airflow/blob/master/airflow/models/taskinstance.py#L1309) code, you can check that by removing the try-catch block exception raised by timeout class will now be handled by the taskinstance instead of the batch_operator class. > > ``` > try: > with timeout(task_copy.execution_timeout.total_seconds()): > result = task_copy.execute(context=context) > except AirflowTaskTimeout: > task_copy.on_kill() > raise > ``` Yeah, I know that. But I was wondering where the "AirflowTaskTimeout" is thrown. the AWS provider is well.. a bit convoluted ... So I guess (purely because this is the only place where the exception is thrown in AWS provider) it is the "wait_for_task_execution" in AWSDataSycnHook that throws it. If so - can you please: a) add some comments describing what's going on - mention that the method might throw the timeout exception and that it should be passed back to execute method (this way the next person coming here will not have to dig deeper and will not add the try/catch back. b) can you please add a unit test case with this case to prevent regressions? ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected]
