[
https://issues.apache.org/jira/browse/MAPREDUCE-6640?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Robert Kanter updated MAPREDUCE-6640:
-------------------------------------
Attachment: MAPREDUCE-6640.001.patch
The patch makes it so that instead of {{<jobHistoryFile>}}, it's now
{{<jobHistoryFile|jobId>}}. When given a Job ID, it will fetch the Job's info
and use that to look up the jhist file location, and then use the old existing
behavior on that file. Besides the obvious changes required for that, I also
had to fix the broken API that should have returned the jhist file location,
which wasn't hooked up before.
The patch is somewhat large because it deletes a jhist file that's no longer
needed by any unit tests.
> mapred job -history command should be able to take Job ID
> ---------------------------------------------------------
>
> Key: MAPREDUCE-6640
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-6640
> Project: Hadoop Map/Reduce
> Issue Type: Improvement
> Components: client
> Reporter: Robert Kanter
> Assignee: Robert Kanter
> Attachments: MAPREDUCE-6640.001.patch
>
>
> The {{mapred job -history}} command currently accepts only a
> {{jobHistoryFile>}}. It would be much easier for the user if it could also
> accept a Job ID so the user doesn't have to go and find the jhist file. It
> would also be more consistent with the other commands, which generally take a
> Job ID.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
