[
https://issues.apache.org/jira/browse/YARN-5889?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15732320#comment-15732320
]
Eric Payne commented on YARN-5889:
----------------------------------
Thanks [~leftnoteasy]. I am still absorbing your previous comment, but I have
couple of thoughts.
bq. That is why we need an async thread to do all these stuffs: we cannot put a
computation which is O(N) to allocation thread.
If all apps are asking for resources, then the current algorithm already does
this calculation in the allocation thread.
bq. This computation is more expensive, it should be O(N), N is number of
applications in the queue.
I believe that N is the number of users, not applications. So, theoretically,
the calculations could be based on each user's {{used}} value. Is that accurate?
> Improve user-limit calculation in capacity scheduler
> ----------------------------------------------------
>
> Key: YARN-5889
> URL: https://issues.apache.org/jira/browse/YARN-5889
> Project: Hadoop YARN
> Issue Type: Bug
> Components: capacity scheduler
> Reporter: Sunil G
> Assignee: Sunil G
> Attachments: YARN-5889.v0.patch, YARN-5889.v1.patch,
> YARN-5889.v2.patch
>
>
> Currently user-limit is computed during every heartbeat allocation cycle with
> a write lock. To improve performance, this tickets is focussing on moving
> user-limit calculation out of heartbeat allocation flow.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
