[ http://issues.apache.org/jira/browse/HADOOP-484?page=all ]

Owen O'Malley resolved HADOOP-484.
----------------------------------

    Resolution: Duplicate

Splitting a reduce would be difficult and error-prone while it was in progress. 
The standard approach for this problem is to use speculative execution to 
shorten the tail. It seems to be very effective.

> Additional splilts for last reduces?
> ------------------------------------
>
>                 Key: HADOOP-484
>                 URL: http://issues.apache.org/jira/browse/HADOOP-484
>             Project: Hadoop
>          Issue Type: Improvement
>          Components: mapred
>            Reporter: arkady borkovsky
>
> Often last few reduces take very long.  
> Would it make sense, if hardware is available, to resplit their inputs into 
> smaller chunks and to run multiple task instead?

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: 
http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Reply via email to