hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Arun C Murthy (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-4642) MiniMRClientClusterFactory should not use job.setJar()
Date Mon, 10 Sep 2012 13:14:07 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-4642?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13451948#comment-13451948
] 

Arun C Murthy commented on MAPREDUCE-4642:
------------------------------------------

Sorry to come in late. 

I'm confused - setJar is a standard user-api - this seems like a bug in Oozie if it can't
deal with it?

In the worst case - we can fix MR1 to use setJar too, but this fix seems off?
                
> MiniMRClientClusterFactory should not use job.setJar()
> ------------------------------------------------------
>
>                 Key: MAPREDUCE-4642
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-4642
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: test
>            Reporter: Robert Kanter
>            Assignee: Robert Kanter
>             Fix For: 2.0.2-alpha
>
>         Attachments: MAPREDUCE-4642.patch, MAPREDUCE-4642.patch
>
>
> Currently, MiniMRClientClusterFactory does {{job.setJar(callerJar)}} so that the {{callerJar}}
is added to the cache in MR2.  However, this makes the resulting configuration inconsistent
between MR1 and MR2 as in MR1 the job jar is not set and in MR2 its set to the {{callerJar}}.
 This difference can also cause some tests to fail in Oozie.  We should instead use the {{job.addCacheFile()}}
method.  

--
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