[
https://issues.apache.org/jira/browse/YARN-11792?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
zeekling updated YARN-11792:
----------------------------
Priority: Critical (was: Major)
> we need kill am container in the case of work-preserving AM restart
> -------------------------------------------------------------------
>
> Key: YARN-11792
> URL: https://issues.apache.org/jira/browse/YARN-11792
> Project: Hadoop YARN
> Issue Type: Bug
> Reporter: zeekling
> Priority: Critical
> Attachments: image-2025-03-07-09-56-17-966.png,
> image-2025-03-07-09-56-52-821.png, image-2025-03-07-09-59-05-265.png
>
>
> When the NM is powered off unexpectedly, the previous AM is not killed. As a
> result, the available NM resources are negative.
>
> !image-2025-03-07-09-56-52-821.png!
> old code :
> !image-2025-03-07-09-56-17-966.png!
>
>
> I think am should be killed at this time.
>
> like this :
> !image-2025-03-07-09-59-05-265.png!
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
