AngersZhuuuu commented on pull request #33780:
URL: https://github.com/apache/spark/pull/33780#issuecomment-903740725


   > the main thing here is to make sure we don't report fail when we aren't 
sure. I need to remember all the cases. I assume in this case you are running 
client batch mode? ie not pyspark or spark-shell
   
   Since for yarn-client mode.
   1.  there are some other component integration with spark use this exit code 
to judge if the job running well, the exit code need to be clear.
   2. if we call sc.stop() in program's end or trigger by ShutfownHook, in 
driver side, it's no matter if MonitorThread get reply and call `sc.stop()`
   3. If it's a good network env, AM should receive the Shutdown message and 
finish with SUCCESS.
   4. If's a bad network, AM can't connect with driver, it always be a failed 
application, but know, it return SUCCESS.
   


-- 
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]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to