[
https://issues.apache.org/jira/browse/MAPREDUCE-3027?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13107449#comment-13107449
]
Vinod Kumar Vavilapalli commented on MAPREDUCE-3027:
----------------------------------------------------
Oh yeah, this must've been a mistake in MAPREDUCE-2896. We should bump
YarnConfiguration.ABORTED_CONTAINER_EXIT_STATUS to -2000 or something like
that. I also see another ref to "-100" in ContainerStatus+75, can you fix that
too?
> MR-279: Completed container exit state needs to be enhanced to differentiate
> between container aborts/failures and actual application process exit codes
> --------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: MAPREDUCE-3027
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3027
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2
> Affects Versions: 0.23.0
> Reporter: Hitesh Shah
> Assignee: Hitesh Shah
> Fix For: 0.23.0
>
>
> Currently, a completed container's exit status is set to -100 to denote the
> container being killed by the framework either as a result of the application
> releasing the container or a node failure. An application process may also
> return an exit code of -100 creating an ambiguity.
>
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
