[
https://issues.apache.org/jira/browse/YARN-914?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16256106#comment-16256106
]
Junping Du commented on YARN-914:
---------------------------------
Client side graceful decommission work has been done together with proper
document, so we should claim part of goal is achieved. I think we should
separate server side decommission work into phase two with fixing HA issues,
Jason Format issues, and other enhancements, which is helpful to make list
cleaner. If nobody against, I will create a new Umbrella jira (and new branch)
for moving all open JIRAs to under that one.
> (Umbrella) Support graceful decommission of nodemanager
> -------------------------------------------------------
>
> Key: YARN-914
> URL: https://issues.apache.org/jira/browse/YARN-914
> Project: Hadoop YARN
> Issue Type: New Feature
> Components: graceful
> Affects Versions: 2.0.4-alpha
> Reporter: Luke Lu
> Assignee: Junping Du
> Attachments: Gracefully Decommission of NodeManager (v1).pdf,
> Gracefully Decommission of NodeManager (v2).pdf,
> GracefullyDecommissionofNodeManagerv3.pdf
>
>
> When NMs are decommissioned for non-fault reasons (capacity change etc.),
> it's desirable to minimize the impact to running applications.
> Currently if a NM is decommissioned, all running containers on the NM need to
> be rescheduled on other NMs. Further more, for finished map tasks, if their
> map output are not fetched by the reducers of the job, these map tasks will
> need to be rerun as well.
> We propose to introduce a mechanism to optionally gracefully decommission a
> node manager.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
