sandflee commented on YARN-4495:

to [~mding],  
1,  we have a StateMachine in AM to track every instance(corresponding to a 
container), if we updating container resource, it will come to UPDATING state, 
and waiting for a relpy to back to RUNNING state.
2,  containerId is essential, we must remove corresponding containerId from 
pendingChange to avoid resend this containerId request.

to [~wangda],
ok, I'll have a test


> add a way to tell AM container increase/decrease request is invalid
> -------------------------------------------------------------------
>                 Key: YARN-4495
>                 URL: https://issues.apache.org/jira/browse/YARN-4495
>             Project: Hadoop YARN
>          Issue Type: Improvement
>            Reporter: sandflee
>         Attachments: YARN-4495.01.patch
> now RM may pass InvalidResourceRequestException to AM or just ignore the 
> change request, the former will cause AMRMClientAsync down. and the latter 
> will leave AM waiting for the relay.  

This message was sent by Atlassian JIRA

Reply via email to