hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hemanth Yamijala (JIRA)" <j...@apache.org>
Subject [jira] Commented: (MAPREDUCE-1451) Minor lock inversions in capacity scheduler
Date Thu, 04 Feb 2010 03:38:27 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-1451?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12829420#action_12829420
] 

Hemanth Yamijala commented on MAPREDUCE-1451:
---------------------------------------------

Thanks for reporting this, Todd. Seems related to startup and refresh happening at the same
time. The third cycle is in a test case. Good to investigate and fix nevertheless.

> Minor lock inversions in capacity scheduler
> -------------------------------------------
>
>                 Key: MAPREDUCE-1451
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1451
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: contrib/capacity-sched
>    Affects Versions: 0.22.0
>            Reporter: Todd Lipcon
>         Attachments: capacity-lock-inversion-2.png, capacity-lock-inversion-3.png, capacity-lock-inversion.png
>
>
> Running jcarder on the contrib unit tests yielded 3 lock inversions in the capacity scheduler.
I don't know enough about the scheduler to know if they're problems - they look harmless enough
since they seem to all be concerned with the startup code, which probably doesn't run concurrently
with other stuff. Attaching the images of the cycles here.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message