[
https://issues.apache.org/jira/browse/YUNIKORN-1049?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17500430#comment-17500430
]
Wilfred Spiegelenburg commented on YUNIKORN-1049:
-------------------------------------------------
[~pbacsko] I think we need to unlink YUNIKORN-1053 from this umbrella and
handle it as a separate improvement.
> [Umbrella] Enhance Yunikorn metrics
> -----------------------------------
>
> Key: YUNIKORN-1049
> URL: https://issues.apache.org/jira/browse/YUNIKORN-1049
> Project: Apache YuniKorn
> Issue Type: Improvement
> Components: metrics
> Reporter: Peter Bacsko
> Assignee: Peter Bacsko
> Priority: Major
>
> From the point of view of supportability, it's important to expose more data
> and statistics about Yunikorn to aid debugging and troubleshooting.
> Data to collect:
> * Go Runtime metrics
> * Cluster metrics
> * Node metrics
> * Queue metrics
> It's also worth comparing current metrics to what have been implemented in
> Hadoop YARN over the years. Although YARN is more than just a scheduler, it
> can still give us insights about what kind of data we can consider to collect.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
