[ https://issues.apache.org/jira/browse/MAPREDUCE-1807?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12869864#action_12869864
]
Sreekanth Ramakrishnan commented on MAPREDUCE-1807:
---------------------------------------------------
Currently in trunk and .21 of hadoop, we have removed all the {{MiniMRCluster}} test to a
new test case and {{TestQueueManager}} is plain unit test. All the integration tests have
been moved to {{TestQueueManagerWithJobTracker}} which has only 3 test cases and all of the
same runs under a minute even. Currently in latest build the same took only 30 seconds.
This was done as part of MAPREDUCE-28.
Isn't solution at MAPREDUCE-28 sufficient?
> TestQueueManager can take long enough to time out
> -------------------------------------------------
>
> Key: MAPREDUCE-1807
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-1807
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Reporter: Dick King
>
> Sometimes TestQueueManager takes such a long time that the JUnit engine times out and
declares it a failure. We should fix this, possibly by splitting the file's 19 test cases
into two or more manageable test sets.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
|