You are viewing a plain text version of this content. The canonical link for it is here.

Posted to yarn-issues@hadoop.apache.org by "Robert Joseph Evans (JIRA)" <ji...@apache.org> on 2013/03/04 17:17:13 UTC

[jira] [Commented] (YARN-448) Remove unnecessary hflush from log aggregation

    [ https://issues.apache.org/jira/browse/YARN-448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13592321#comment-13592321 ] 

Robert Joseph Evans commented on YARN-448:
------------------------------------------

Yes I don't see any real reason to have that hflush there. If log aggregation crashed either before or after that point it really would not make that much of a difference. +1, assuming Jenkins comes back with a +1.
                
> Remove unnecessary hflush from log aggregation
> ----------------------------------------------
>
>                 Key: YARN-448
>                 URL: https://issues.apache.org/jira/browse/YARN-448
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: nodemanager
>    Affects Versions: 0.23.7, 2.0.4-beta
>            Reporter: Kihwal Lee
>            Assignee: Kihwal Lee
>         Attachments: yarn-448.patch.txt
>
>
> AggregatedLogFormat#writeVersion() calls hflush() after writing the version. Calling hflush does not seem to be necessary. It can add a lot of load to hdfs in a big busy cluster.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira