You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@tez.apache.org by "Siddharth Seth (JIRA)" <ji...@apache.org> on 2013/12/17 23:16:08 UTC

[jira] [Updated] (TEZ-681) Grouping generates incorrect splits if multiple DNs run on a single node

     [ https://issues.apache.org/jira/browse/TEZ-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Siddharth Seth updated TEZ-681:
-------------------------------

    Attachment: grouping_tmp_patch.txt

Temporary patch to verify the issue - with lots of comments and some logging. May or may not be what a final fix looks like. [~bikassaha], do you want to look at this, otherwise I'll take it up.

> Grouping generates incorrect splits if multiple DNs run on a single node
> ------------------------------------------------------------------------
>
>                 Key: TEZ-681
>                 URL: https://issues.apache.org/jira/browse/TEZ-681
>             Project: Apache Tez
>          Issue Type: Bug
>            Reporter: Siddharth Seth
>            Priority: Critical
>         Attachments: grouping_tmp_patch.txt
>
>
> When running multiple DNs on a single node (MiniDFSCluster for example), grouping ends up using the same split multiple times while grouping - so the end result is Grouped splits with some original splits repeated, and others completely missed out.



--
This message was sent by Atlassian JIRA
(v6.1.4#6159)