[
https://issues.apache.org/jira/browse/HADOOP-4927?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12670028#action_12670028
]
Doug Cutting commented on HADOOP-4927:
--------------------------------------
> Unless there's a non-FileOutputFormat use case [ ... ]
I see Chris's point and agree. Unless there's a strong reason to put features
in the kernel we should prefer to put them in library code, keeping the kernel
minimal. Are there non-FileInputFormats that need this feature?
A wrapper implementation is a bit harder to use, since folks would need to both
set the job's outputformat to the wrapper, and set the wrapper's parameter to
the real output format: two changes instead of just setting a single parameter,
although it is more generic. We could perhaps implement both: a flag for
FileOutputFormat and a wrapper OutputFormat for folks who've not subclassed
FileOutputFormat?
> Part files on the output filesystem are created irrespective of whether the
> corresponding task has anything to write there
> --------------------------------------------------------------------------------------------------------------------------
>
> Key: HADOOP-4927
> URL: https://issues.apache.org/jira/browse/HADOOP-4927
> Project: Hadoop Core
> Issue Type: New Feature
> Components: mapred
> Reporter: Devaraj Das
> Assignee: Jothi Padmanabhan
> Fix For: 0.21.0
>
> Attachments: hadoop-4927-v1.patch, hadoop-4927-v2.patch,
> hadoop-4927.patch
>
>
> When OutputFormat.getRecordWriter is invoked, a part file is created on the
> output filesystem. But the created RecordWriter is not used until the
> OutputCollector.collect call is made by the task (user's code). This results
> in empty part files even if the OutputCollector.collect is never invoked by
> the corresponding tasks.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
