hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Amar Kamat (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-2141) speculative execution start up condition based on completion time
Date Fri, 08 Feb 2008 15:50:07 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-2141?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12567071#action_12567071
] 

Amar Kamat commented on HADOOP-2141:
------------------------------------

Dumping some logs from my recent runs
{noformat}
2008-02-08 15:09:19,459 INFO org.apache.hadoop.mapred.TaskInProgress: Error from task_200802080908_0005_r_001459_0:
Task task_200802080908_0005_r_001459_0 failed to report status for 605 seconds. Killing!
2008-02-08 15:09:19,460 INFO org.apache.hadoop.mapred.JobTracker: Removed completed task 'task_200802080908_0005_r_001459_0'
from 'tracker_gs205019.inktomisearch.com:gs205019.inktomisearch.com/76.13.184.103:58495'
2008-02-08 15:09:19,474 INFO org.apache.hadoop.mapred.TaskRunner: Discarded output of task
'task_200802080908_0005_r_001459_0' - hdfs://gs205514.inktomisearch.com:57972/user/amarrk/output/_task_200802080908_0005_r_001459_0
2008-02-08 15:09:19,513 INFO org.apache.hadoop.mapred.JobInProgress: Choosing normal task
tip_200802080908_0005_r_001459
2008-02-08 15:09:19,514 INFO org.apache.hadoop.mapred.JobTracker: Adding task 'task_200802080908_0005_r_001459_1'
to tip tip_200802080908_0005_r_001459, for tracker 'tracker_gs205440.inktomisearch.com:gs205440.inktomisearch.com/76.13.187.49:55843'
2008-02-08 15:09:19,517 INFO org.apache.hadoop.mapred.JobInProgress: Choosing speculative
task tip_200802080908_0005_r_001459
2008-02-08 15:09:19,517 INFO org.apache.hadoop.mapred.JobTracker: Adding task 'task_200802080908_0005_r_001459_2'
to tip tip_200802080908_0005_r_001459, for tracker 'tracker_gs205190.inktomisearch.com:gs205190.inktomisearch.com/76.13.185.109:58837'
{noformat}
The main task and speculative one got executed back to back.

> speculative execution start up condition based on completion time
> -----------------------------------------------------------------
>
>                 Key: HADOOP-2141
>                 URL: https://issues.apache.org/jira/browse/HADOOP-2141
>             Project: Hadoop Core
>          Issue Type: Improvement
>          Components: mapred
>    Affects Versions: 0.15.0
>            Reporter: Koji Noguchi
>            Assignee: Arun C Murthy
>             Fix For: 0.17.0
>
>
> We had one job with speculative execution hang.
> 4 reduce tasks were stuck with 95% completion because of a bad disk. 
> Devaraj pointed out 
> bq . One of the conditions that must be met for launching a speculative instance of a
task is that it must be at least 20% behind the average progress, and this is not true here.
> It would be nice if speculative execution also starts up when tasks stop making progress.
> Devaraj suggested 
> bq. Maybe, we should introduce a condition for average completion time for tasks in the
speculative execution check. 

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message