[
https://issues.apache.org/jira/browse/YARN-5566?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15439437#comment-15439437
]
Robert Kanter commented on YARN-5566:
-------------------------------------
I've been super busy and haven't had a chance to look at YARN-4676 again. I've
actually been testing the client-side graceful decom without YARN-4676 applied,
so the logic that was removed by it is actually still there.
I agree that UNHEALTHY isn't a state that should occur if the initial state is
DECOMMISSIONING; as you pointed out, the code doesn't go to UNHEALTHY from
DECOMMISSIONG. That's partly why I was unsure why my patch fixed the problem;
the node never goes to UNHEALTHY, so it doesn't make sense why that would help.
I'll try to dig into this some more.
> client-side NM graceful decom doesn't trigger when jobs finish
> --------------------------------------------------------------
>
> Key: YARN-5566
> URL: https://issues.apache.org/jira/browse/YARN-5566
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: nodemanager
> Affects Versions: 2.8.0
> Reporter: Robert Kanter
> Assignee: Robert Kanter
> Attachments: YARN-5566.001.patch
>
>
> I was testing the client-side NM graceful decommission and noticed that it
> was always waiting for the timeout, even if all jobs running on that node (or
> even the cluster) had already finished.
> For example:
> # JobA is running with at least one container on NodeA
> # User runs client-side decom on NodeA at 5:00am with a timeout of 3 hours
> --> NodeA enters DECOMMISSIONING state
> # JobA finishes at 6:00am and there are no other jobs running on NodeA
> # User's client reaches the timeout at 8:00am, and forcibly decommissions
> NodeA
> NodeA should have decommissioned at 6:00am.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
