[
https://issues.apache.org/jira/browse/MAPREDUCE-4292?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13285819#comment-13285819
]
Jason Lowe commented on MAPREDUCE-4292:
---------------------------------------
I tried reproducing this with 2.0.0-alpha on a single-node cluster by modifying
wordcount to fail map IDs < 4. As expected, the job failed soon after a map
failed four attempts. Explicitly setting mapred.reduce.tasks=1 for the job had
no effect. Could you attach the job config? I'm wondering if there are other
property settings that are affecting the behavior of the job. For example, is
mapreduce.map.failures.maxpercent set to a non-zero value?
> Job is hanging forever when some maps are failing always
> --------------------------------------------------------
>
> Key: MAPREDUCE-4292
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4292
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2
> Affects Versions: 2.0.0-alpha
> Reporter: Nishan Shetty
> Priority: Critical
> Attachments: syslog
>
>
> Set property "mapred.reduce.tasks" to some value greater than zero
> I have a job in which some maps are failing always.
> Observations:
> 1.Map phase is completing with 100%(with succeeded and failed maps).
> 2.Reduce phase is not progressing further after 32%.
> 3.After map phase is completed job is hanging forever.
> Expected that job should be failed after waiting for some time.
--
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
