hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Robert Joseph Evans (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-4229) Intern counter names in the JT
Date Wed, 17 Oct 2012 20:00:03 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-4229?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13478314#comment-13478314

Robert Joseph Evans commented on MAPREDUCE-4229:

Sorry it has taken me so long to respond. I have been a bit swamped lately.  The new patch
looks really good.  It is simple and looks like it could help a lot with the memory usage.
 Do you have any actual heap comparisons that you can show us?  I just have a difficult time
checking in a "performance" fix without some test, manual or otherwise to show the impact
it is having and if there is still more that could be done in a follow up JIRA.  I know that
YourKit profiler has some nice Heap Dump analysis to look for duplicate strings. If you have
some numbers ready that would be great otherwise I will try and find some time this week to
see if I can come up with anything.
> Intern counter names in the JT
> ------------------------------
>                 Key: MAPREDUCE-4229
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-4229
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>          Components: jobtracker
>    Affects Versions: 1.0.2, 3.0.0, 2.0.2-alpha
>            Reporter: Todd Lipcon
>         Attachments: MAPREDUCE-4229-branch-0.23.patch, MAPREDUCE-4229.patch
> In our experience, most of the memory in production JTs goes to storing counter names
(String objects and character arrays). Since most counter names are reused again and again,
it would be a big memory savings to keep a hash set of already-used counter names within a
job, and refer to the same object from all tasks.

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

View raw message