[jira] [Commented] (MAPREDUCE-4749) Killing multiple attempts of a task taker longer as more attempts are killed

["Vinod Kumar Vavilapalli (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/MAPREDUCE-4749?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13487239#comment-13487239 ] 

Vinod Kumar Vavilapalli commented on MAPREDUCE-4749:
----------------------------------------------------

Added Bobby and Ravi onto the watch list as they are linked to MAPREDUCE-4088.

So what is happening is that because of the patch, every clean up action like KillTaskAction, KillJobAction *ALWAYS* wait for 5 seconds before processing. Because of this, all clean up actions delay, and because till TT actually cleans up a task/job, JobTracker keeps on dishing out new duplicate clean-up actions for the same task/job. This causes a avalanche of clean-up actions, even only one task is cleanup, the queue grows continuously and so later, even killing jobs will take for ever (even though clients are not blocked).
                
> Killing multiple attempts of a task taker longer as more attempts are killed
> ----------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-4749
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-4749
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>    Affects Versions: 1.1.0
>            Reporter: Arpit Gupta
>            Assignee: Vinod Kumar Vavilapalli
>
> The following was noticed on a mr job running on hadoop 1.1.0
> 1. Start an mr job with 1 mapper
> 2. Wait for a min
> 3. Kill the first attempt of the mapper and then subsequently kill the other 3 attempts in order to fail the job
> The time taken to kill the task grew exponentially.
> 1st attempt was killed immediately.
> 2nd attempt took a little over a min
> 3rd attempt took approx. 20 mins
> 4th attempt took around 3 hrs.
> The command used to kill the attempt was "hadoop job -fail-task"
> Note that the command returned immediately as soon as the fail attempt was accepted but the time the attempt was actually killed was as stated above.

--
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