You are viewing a plain text version of this content. The canonical link for it is here.
Posted to mapreduce-issues@hadoop.apache.org by "Tom White (JIRA)" <ji...@apache.org> on 2011/03/17 06:24:29 UTC
[jira] Updated: (MAPREDUCE-2392) TaskTracker shutdown in the tests
sometimes take 60s
[ https://issues.apache.org/jira/browse/MAPREDUCE-2392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Tom White updated MAPREDUCE-2392:
---------------------------------
Attachment: MAPREDUCE-2392.patch
This is reminiscent of HADOOP-5380, except in this case it is the TT-JT communication that is timing out.
The problem is that calling interrupt() on the TaskTracker thread can (by chance) cause it to interrupt the heartbeat RPC call, which then takes 60 seconds to timeout ("ipc.ping.interval") from readInt() in org.apache.hadoop.ipc.Client$Connection.receiveResponse.
This can be fixed by removing the interrupt call to the TaskTracker, since we already call shutdown() on the TaskTracker then join() on the thread running it.
> TaskTracker shutdown in the tests sometimes take 60s
> ----------------------------------------------------
>
> Key: MAPREDUCE-2392
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-2392
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Reporter: Tom White
> Assignee: Tom White
> Attachments: MAPREDUCE-2392.patch
>
>
> There are a lot of the following in the test logs:
> {noformat}
> 2011-03-16 13:47:02,267 INFO mapred.TaskTracker (TaskTracker.java:shutdown(1275)) - Shutting down StatusHttpServer
> 2011-03-16 13:48:02,349 ERROR mapred.TaskTracker (TaskTracker.java:offerService(1609)) - Caught exception: java.io.IOException: Call to localhost/127.0.0.1:57512 failed on local exception: java.nio.channels.ClosedByInterruptException
> {noformat}
> Note there is over one minute between the first line and the second.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira