[
https://issues.apache.org/jira/browse/MAPREDUCE-5841?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13976992#comment-13976992
]
Jason Lowe commented on MAPREDUCE-5841:
---------------------------------------
bq. Anyhow, the reason I arrived at the current fix is that once the state
transition is completed, LocalContainerLauncher.stop() is invoked, and that
shuts down the ExecutorService for the task runner. And that shut down
interrupts any running task. So the end result is pretty much the same.
Yes I understand how the end result is the same for the mapred job -kill case,
but doesn't the job still hang in the mapred job -fail-task or mapred job
-kill-task case if the task is stuck? Since we fake the exit of the
"container" the AM will attempt to re-launch the task, the new task attempt
will get submitted to the executor queue, but I think it hangs in the queue
waiting for the stuck task since it's only executing one task at a time. At
that point the only recourse is to kill the entire job.
If you could poke at it a bit more I think it'd be good to understand what's
going on if we cancel the task. If it proves particularly sticky then we can
consider putting this in and filing a followup JIRA to track the remaining
fail-task/kill-task issue.
> uber job doesn't terminate on getting mapred job kill
> -----------------------------------------------------
>
> Key: MAPREDUCE-5841
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-5841
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2
> Affects Versions: 2.3.0
> Reporter: Sangjin Lee
> Assignee: Sangjin Lee
> Attachments: mapreduce-5841.patch
>
>
> If you issue a "mapred job -kill" against a uberized job, the job (and the
> yarn application) state transitions to KILLED, but the application master
> process continues to run. The job actually runs to completion despite the
> killed status.
> This can be easily reproduced by running a sleep job:
> {noformat}
> hadoop jar hadoop-mapreduce-client-jobclient-2.3.0-tests.jar sleep -m 1 -r 0
> -mt 300000
> {noformat}
> Issue a kill with "mapred job -kill \[job-id\]". The UI will show the job
> (app) is in the KILLED state. However, you can see the application master is
> still running.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
