[
https://issues.apache.org/jira/browse/MAPREDUCE-2705?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13068600#comment-13068600
]
Thomas Graves commented on MAPREDUCE-2705:
------------------------------------------
The solution implemented is to have it start a new thread to do the actual
localization and task launching for that particular task, that way it doesn't
block other tasks behind it if the localization of one task is taking a long
time.
> tasks localized and launched serially by TaskLauncher - causing other tasks
> to be delayed
> -----------------------------------------------------------------------------------------
>
> Key: MAPREDUCE-2705
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-2705
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: tasktracker
> Affects Versions: 0.20.205.0, 0.23.0
> Reporter: Thomas Graves
> Assignee: Thomas Graves
> Fix For: 0.20.205.0, 0.23.0
>
> Attachments: MAPREDUCE-2705-branch20.patch
>
>
> The current TaskLauncher serially launches new tasks one at a time. During
> the launch it does the localization and then starts the map/reduce task.
> This can cause any other tasks to be blocked waiting for the current task to
> be localized and started. In some instances we have seen a task that has a
> large file to localize (1.2MB) block another task for about 40 minutes. This
> particular task being blocked was a cleanup task which caused the job to be
> delayed finishing for the 40 minutes.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
