TisonKun edited a comment on issue #9719: [FLINK-14010][coordination] Shutdown 
cluster if ResourceManager terminated unexpectedly
URL: https://github.com/apache/flink/pull/9719#issuecomment-534370206
 
 
   Hi @tillrohrmann Thanks for your review.
   
   I add a `DispatcherResourceManagerComponentTest` and expose some visibility, 
please double check it doesn't cause other issues.
   
   Besides, I find another code change required by this approach that actually 
we are unable to `deregisterApplication` on `DispatcherResourceManager` closed 
because RM is already terminated.
   
   It seems we twist codes a bit more so although the logic is correct at the 
moment I still prefer just call `onFatalError` in 
`YarnResourceManager#onShutdownRequest` which fixes the original issue 
specifically. We cannot do anything when we lost connection to YARN RM and lost 
leadership semantically. Anything I miss we gain by gracefully closing 
ClusterEntrypoint?

----------------------------------------------------------------
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]


With regards,
Apache Git Services

Reply via email to