[
https://issues.apache.org/jira/browse/HADOOP-939?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12517672
]
Joydeep Sen Sarma commented on HADOOP-939:
------------------------------------------
am a new hadoop user - i am looking into how much of a warehouse type
infrastructure can be implemented over hadoop. in some cases - i would like to
have the flexibility of having the output partitioned by the mapoutput key -
but i am not interested in it being sorted. the directmapoutputcollector takes
away sorting - but it also takes away partitioning. a lighter hammer would be
useful imho.
> No-sort optimization
> --------------------
>
> Key: HADOOP-939
> URL: https://issues.apache.org/jira/browse/HADOOP-939
> Project: Hadoop
> Issue Type: New Feature
> Components: mapred
> Environment: all
> Reporter: Doug Judd
>
> There should be a way to tell the mapred framework that the output of the
> map() phase will already be sorted. The Reduce phase can just merge the
> intermediate files together without sorting.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
