[
https://issues.apache.org/jira/browse/YARN-6677?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16493926#comment-16493926
]
Haibo Chen commented on YARN-6677:
----------------------------------
Thanks [[email protected]] for the comment.
{quote}Could you just update DefaultOOMHandler with your code?
{quote}
This is doable, but will change the current DefaultOOMHandler behavior
drastically. Specifically, in the case of OPPORTUNISTIC containers, they will
be killed regardless of whether their individual usage is over their limit or
not. We can preserve the behavior for GUARANTEED containers, without the two
level sort comparator, as well. Let me know if that's okay with you.
> Preempt opportunistic containers when root container cgroup goes over memory
> limit
> ----------------------------------------------------------------------------------
>
> Key: YARN-6677
> URL: https://issues.apache.org/jira/browse/YARN-6677
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: nodemanager
> Affects Versions: 3.0.0-alpha3
> Reporter: Haibo Chen
> Assignee: Miklos Szegedi
> Priority: Major
> Attachments: YARN-6677.00.patch
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
