hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Devaraj Das (JIRA)" <j...@apache.org>
Subject [jira] Updated: (HADOOP-968) Reduce shuffle and merge should be done a child JVM
Date Tue, 17 Apr 2007 19:24:15 GMT

     [ https://issues.apache.org/jira/browse/HADOOP-968?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

Devaraj Das updated HADOOP-968:

    Attachment: 968-with-metrics-fix.new.patch

Ok, changed the Timer constructor to the two-argument one. Also, retained the metricsContext.close()
call that I had in the last patch. In general, I think that it is a good idea to call close()
on an object if the close() method is clearly documented. Does this seem all right?

> Reduce shuffle and merge should be done a child JVM
> ---------------------------------------------------
>                 Key: HADOOP-968
>                 URL: https://issues.apache.org/jira/browse/HADOOP-968
>             Project: Hadoop
>          Issue Type: Improvement
>          Components: mapred
>    Affects Versions: 0.10.1
>            Reporter: Owen O'Malley
>         Assigned To: Devaraj Das
>             Fix For: 0.13.0
>         Attachments: 968-reindent.patch, 968-with-metrics-fix.new.patch, 968-with-metrics-fix.patch,
968.apr06.patch, 968.apr10.patch, 968.apr14.patch, 968.apr14.patch, 968.patch
> The Reduce's shuffle and initial merge is done in the TaskTracker's JVM. It would be
better to have it run in the Task's child JVM. The advantages are:
>   1. The class path and environment would be set up correctly.
>   2. User code doesn't need to be loaded into the TaskTracker.
>   3. Lower memory usage and contention in the TaskTracker.

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

View raw message