potiuk commented on issue #26354:
URL: https://github.com/apache/airflow/issues/26354#issuecomment-1250393587

   I believe You need to handle it in your task pod - to check if the other 
pods are failing as expected, and if they are not, the task should fail. This 
is very similar story as with "init" containers monitoring - we do not handle 
currently "automated" closing of such init containers automatically and the 
task should close them as needed (so no daemon kind of tasks).
   
   I think there are far too many cases to handle them automatically. For 
example there are many cases where some containers in your pod will fail 
initially (for example when there is a database not started/initialized yet) 
and the containers will get automatically restarted and will eventually 
succeed. While this is not a "perfect" patterm, it does happen more often than 
not in K8S world. This opens all the kind of problems like "should we wait for 
the containers"? How long? How many retries? etc. etc. 
   
   However maybe others have a different opinion (@dstandish @jedcunningham 
@ephraimbuddy ?) . I will convert it into discussion in case you @karoldob want 
also discuss it. Maybe a new feature will be born out  of it, but for sure, 
this is not a bug.


-- 
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]

Reply via email to