[
https://issues.apache.org/jira/browse/YUNIKORN-2406?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17817401#comment-17817401
]
Craig Condit commented on YUNIKORN-2406:
----------------------------------------
Merged core-side changes, shim PR needs to be committed as well to avoid
breakage as the core logic isn't handled properly by the old code.
> Recovered tasks do not move to Bound state
> ------------------------------------------
>
> Key: YUNIKORN-2406
> URL: https://issues.apache.org/jira/browse/YUNIKORN-2406
> Project: Apache YuniKorn
> Issue Type: Bug
> Components: core - scheduler, shim - kubernetes
> Reporter: Chia-Ping Tsai
> Assignee: Craig Condit
> Priority: Blocker
> Labels: pull-request-available
>
> The following events will update the count of allocated containers.
> 1. release allocation
> 2. schedule new allocation
> 3. remove a node
> Unfortunately, the event of recovering allocations does not update the count.
> Hence, the restarted scheduler shows zero value of allocated containers even
> though there are a bunch of running containers.
> This bug cause the restful API of showing container history will return zero,
> and so the dashboard of WEB UI shows zero also.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
