rcheatham-q commented on issue #42135:
URL: https://github.com/apache/airflow/issues/42135#issuecomment-2596263452

   Can we reopen this? I just encountered this issue on version 2.10.4 and I'm 
confident I can reliably reproduce it. The high-level steps would be:
   
   1. Deploy Airflow in k8s with a task that runs for a few minutes (it can 
simply sleep) and has at least 1 retry configured. I don't think executor 
matters, it could be Celery or Kubernetes.
   1. While the task is running, forcefully kill the pod running the task 
without a termination grace period
   
      ```
      kubectl delete --grace-period=1 pod/...
      ```
   
   1. Wait for the scheduler to detect the zombie task and trigger a retry
   
   ### Expected behavior:
   The scheduler detects the zombie task and schedules a retry
   
   ### Actual behavior:
   The scheduler detects the zombie task and marks it as failed


-- 
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: commits-unsubscr...@airflow.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org

Reply via email to