hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Owen O'Malley (JIRA)" <j...@apache.org>
Subject [jira] Resolved: (HADOOP-179) task tracker ghosts remain after 10 minutes
Date Tue, 16 May 2006 04:01:06 GMT
     [ http://issues.apache.org/jira/browse/HADOOP-179?page=all ]
     
Owen O'Malley resolved HADOOP-179:
----------------------------------

    Resolution: Fixed

This bug is fixed in the most part by my patch to HADOOP-180, which among other things made
the task trackers choose names for themselves based on the host and port number. Because when
task trackers are reinitialized for missing a heartbeat, they will almost always get the same
name, they will not leave behind "ghosts".

> task tracker ghosts remain after 10 minutes
> -------------------------------------------
>
>          Key: HADOOP-179
>          URL: http://issues.apache.org/jira/browse/HADOOP-179
>      Project: Hadoop
>         Type: Bug

>   Components: mapred
>     Versions: 0.1.1
>     Reporter: Owen O'Malley
>     Assignee: Owen O'Malley
>      Fix For: 0.3

>
> I had a bunch of TaskTrackers time out because they were in the middle of job cleanup
and the JobTracker restarted them by responding to emitHeartbeat with UNKNOWN_TASKTRACKER.
Afterwards, I ended up with both the new and the restarted TaskTrackers on the list:
> node1100_1234, 0-10 seconds since heartbeat
> node1100_4321,  >10,000 seconds since heartbeat

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
   http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
   http://www.atlassian.com/software/jira


Mime
View raw message