[ 
https://issues.apache.org/jira/browse/HADOOP-9?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12928934#action_12928934
 ] 

Paul Baclace commented on HADOOP-9:
-----------------------------------

Ari,

This issue was originally NUTCH-181 before Hadoop was split off.  I wrote a 
patch Dec. 29 2005 and used it at archive.org Jan-Feb 2006.  Looking at my old 
notes, I created this issue on Jan. 11 2006, and prepared the patch on Feb. 28 
2006, but it was either lost in a Jira transition or the attachment somehow 
failed.  

When I looked at your patch yesterday, it was similar enough to what I 
remembered (5 years ago) that I thought it must be a revision of the patch I 
did.  Today I found my source and 2005-2006 work notes and it is clear that you 
implemented the change without seeing mine.

Thanks for doing it in same "roulette-y" spirit of my lost patch! 


> mapred.local.dir  temp dir. space allocation limited by smallest area
> ---------------------------------------------------------------------
>
>                 Key: HADOOP-9
>                 URL: https://issues.apache.org/jira/browse/HADOOP-9
>             Project: Hadoop Common
>          Issue Type: Bug
>         Environment: all
>            Reporter: Paul Baclace
>            Assignee: Ari Rabkin
>            Priority: Minor
>             Fix For: 0.19.0
>
>         Attachments: hadoop9.patch
>
>
> When mapred.local.dir is used to specify multiple  temp dir. areas, space 
> allocation limited by smallest area because the temp dir. selection algorithm 
> is "round robin starting from a randomish point".   When round robin is used 
> with approximately constant sized chunks, the smallest area runs out of space 
> first, and this is a fatal error. 
> Workaround: only list local fs dirs in mapred.local.dir with similarly-sized 
> available areas.
> I wrote a patch to JobConf (currenly being tested) which uses df to check 
> available space (once a minute or less often) and then uses an efficient 
> roulette selection to do allocation weighted by magnitude of available space. 

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

Reply via email to