You are viewing a plain text version of this content. The canonical link for it is here.

Posted to common-dev@hadoop.apache.org by "stack (JIRA)" <ji...@apache.org> on 2009/05/09 07:36:45 UTC

[jira] Updated: (HADOOP-5797) Better distribution in the PerformanceEvaluation MapReduce when rows run to the Billions

     [ https://issues.apache.org/jira/browse/HADOOP-5797?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

stack updated HADOOP-5797:
--------------------------

    Attachment: pe.java

Randomize the way we write the splits.

> Better distribution in the PerformanceEvaluation MapReduce when rows run to the Billions
> ----------------------------------------------------------------------------------------
>
>                 Key: HADOOP-5797
>                 URL: https://issues.apache.org/jira/browse/HADOOP-5797
>             Project: Hadoop Core
>          Issue Type: Bug
>            Reporter: stack
>            Assignee: stack
>             Fix For: 0.20.0
>
>         Attachments: pe.java
>
>
> Performance Evaluation when the number of clients was large (and thus number of rows to input) was clumping.... so we'd beat up on one box then the next.

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