[jira] [Updated] (MAPREDUCE-5000) TaskImpl.getCounters() can return the counters for the wrong task attempt when task is speculating

["Jason Lowe (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/MAPREDUCE-5000?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jason Lowe updated MAPREDUCE-5000:
----------------------------------

    Attachment: MAPREDUCE-5000.patch

Patch to have selectBestAttempt use the successful attempt, if available.
                
> TaskImpl.getCounters() can return the counters for the wrong task attempt when task is speculating
> --------------------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-5000
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5000
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: mr-am
>    Affects Versions: 0.23.6
>            Reporter: Jason Lowe
>            Assignee: Jason Lowe
>            Priority: Critical
>         Attachments: MAPREDUCE-5000.patch
>
>
> When a task is speculating and one attempt completes then sometimes the counters for the wrong attempt are aggregated into the total counters for the job.  The scenario looks like this:
> # Two task attempts are racing, _0 and _1
> # _1 finishes first, causing the task to issue a TA_KILL to attempt _0
> # _0 receives TA_KILL, sets progress to 1.0f and waits for container cleanup
> # if TaskImpl.getCounters() is called now, TaskImpl.selectBestAttempt() can return _0 since it is not quite yet in the KILLED state yet progress is maxed out and no other attempt has more progress.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira