Sunil G commented on YARN-4138:

Thank you very much [~mding] for sharing the update.  I am following YARN-1651 

As updated above, ContainerResourceIncreaseAllocationExpirer has to register 
and monitor for every container which are asked to increase container request 
(storing old resource). One doubt is, if multiple requests comes for same 
container, and that increase token expires, do we need to set back the very 
first container resource. I feel we can reset to the first container increase 
request's start resource unit.

> Roll back container resource allocation after resource increase token expires
> -----------------------------------------------------------------------------
>                 Key: YARN-4138
>                 URL: https://issues.apache.org/jira/browse/YARN-4138
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: api, nodemanager, resourcemanager
>            Reporter: MENG DING
>            Assignee: Sunil G
> In YARN-1651, after container resource increase token expires, the running 
> container is killed.
> This ticket will change the behavior such that when a container resource 
> increase token expires, the resource allocation of the container will be 
> reverted back to the value before the increase.

This message was sent by Atlassian JIRA

Reply via email to