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

Posted to mapreduce-issues@hadoop.apache.org by "Luke Lu (JIRA)" <ji...@apache.org> on 2012/07/07 00:43:34 UTC

[jira] [Updated] (MAPREDUCE-4399) Fix performance regression in shuffle

     [ https://issues.apache.org/jira/browse/MAPREDUCE-4399?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Luke Lu updated MAPREDUCE-4399:
-------------------------------

    Attachment: mapreduce-4399-branch-1.patch

The crux of the problem is jetty 6.1.26's internal buffer management with transfer-encoding = trunked. The issue was also independently discovered and analyzed elsewhere:  http://goo.gl/dtM4D

The patch is trivial.
                
> Fix performance regression in shuffle 
> --------------------------------------
>
>                 Key: MAPREDUCE-4399
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-4399
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: tasktracker
>    Affects Versions: 0.20.203.0, 1.0.3
>            Reporter: Luke Lu
>            Assignee: Luke Lu
>         Attachments: mapreduce-4399-branch-1.patch
>
>
> There is a significant (up to 3x) performance regression in shuffle (vs 0.20.2) in the Hadoop 1.x series. Most noticeable with high-end switches.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira