pingzh commented on issue #23950: URL: https://github.com/apache/airflow/issues/23950#issuecomment-1148122514
> That would imply that the pods were killed though, not the individual tasks that are running there, no? In my case, the worker pods don't have any restarts, which means that the resource usage was not a problem at least for these failures. Am I right in thinking this way? good point on the data on pods restart. if the pod runs out of resources, the pod would be killed. do you see this metric `zombies_killed` https://github.com/apache/airflow/blob/2.2.5/airflow/dag_processing/manager.py#L1135? -- 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]
