hama-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Thomas Jungblut (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HAMA-640) Large graph computing cause out of memory
Date Thu, 13 Sep 2012 11:58:08 GMT

    [ https://issues.apache.org/jira/browse/HAMA-640?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13454841#comment-13454841
] 

Thomas Jungblut commented on HAMA-640:
--------------------------------------

Hi Zhuang, this JIRA is just for development tracking. I would advise you to come to our user
mailing list and explain yourself.
We are aware of the memory hunger and improved in HAMA-596 which will be release with 0.6.0.

At least you have multiple options:
- increase the heap of each task in your job by lifting the -Xmx configuration property in
"bsp.child.java.opts". 
- use the trunk version (build for yourself)
- wait for the next release

I will close this, thanks for your report.
                
> Large graph computing cause out of memory
> -----------------------------------------
>
>                 Key: HAMA-640
>                 URL: https://issues.apache.org/jira/browse/HAMA-640
>             Project: Hama
>          Issue Type: Question
>          Components: graph
>    Affects Versions: 0.5.0
>         Environment: hadoop-1.0.3   hama-0.5.0
>            Reporter: Zhuang Kechen
>
> When I test some small graphs PageRank on distributed environment, everything went all
right. When I upload a Larger graph(758Mb) on HDFS, the graph format is all the same, and
the PageRank went right on SuperStep 0,1,2, then job failed after that, out of memory. I didn't
change the examples, the test use the hama-examples pagerank.
> Somebody please help...Thanks!!

--
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

Mime
View raw message