Mailing-List: contact hadoop-dev-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: hadoop-dev@lucene.apache.org
Message-ID: <26528189.1146714738043.JavaMail.jira@brutus>
Date: Thu, 4 May 2006 03:52:18 +0000 (GMT+00:00)
From: "Owen O'Malley (JIRA)" <jira@apache.org>
To: hadoop-dev@lucene.apache.org
Subject: [jira] Commented: (HADOOP-182) lost task trackers should not update
 status of completed jobs
In-Reply-To: <18534471.1146509326381.JavaMail.jira@brutus>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

    [ http://issues.apache.org/jira/browse/HADOOP-182?page=comments#action_12377692 ] 

Owen O'Malley commented on HADOOP-182:
--------------------------------------

Note that marking completed reduces of running jobs as failed is _really_ bad because if the file already exists, I believe the tip will fail, thereby killing the job.

> lost task trackers should not update status of completed jobs
> -------------------------------------------------------------
>
>          Key: HADOOP-182
>          URL: http://issues.apache.org/jira/browse/HADOOP-182
>      Project: Hadoop
>         Type: Bug

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

>
> When a Task Tracker is lost (by not sending a heartbeat for 10 minutes), the JobTracker marks the tasks that were active on that node as failed. There are two issues:
>    1. No task from a completed or failed job should be modified.
>    2. No reduces should be marked as failed, since their output is in dfs and therefore not only on the dead node.

-- 
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