[
https://issues.apache.org/jira/browse/MAPREDUCE-4389?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13405502#comment-13405502
]
Harsh J commented on MAPREDUCE-4389:
------------------------------------
Output formats always look at OutputK/V classes. The intermediate output writer
always looks at MapOutputK/V classes, which is checked only if numReduces > 0.
Hence, for all final outputs, be it in map stage or be it in reduce stage, its
only the former that matters to the O/F.
I'm for closing this JIRA as invalid, unless you feel a doc improvement at
least needs to be made.
> Map only jobs should work even for MapOutput(Key|Value)Class configuration
> --------------------------------------------------------------------------
>
> Key: MAPREDUCE-4389
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4389
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: job submission, mrv2
> Affects Versions: 2.0.0-alpha, trunk
> Reporter: Bhallamudi Venkata Siva Kamesh
> Assignee: Bhallamudi Venkata Siva Kamesh
>
> I have written a Map only job and have set MapOutput(Key|Value)Class.
> Job was passing when I set TextOutputFormat, however, when I changed the OF
> to SequenceFileOutputFormat, job started failing.
> Later, I have set them to Output(Key|Value)Class, and job was passing.
> I *think*, for Map only jobs, jobs should pass even for
> MapOutput(Key|Value)Class.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
