hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ravi Gummadi (JIRA)" <j...@apache.org>
Subject [jira] Commented: (MAPREDUCE-1829) JobInProgress.findSpeculativeTask should use min() to find the candidate instead of sort()
Date Fri, 11 Jun 2010 05:31:14 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-1829?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12877719#action_12877719
] 

Ravi Gummadi commented on MAPREDUCE-1829:
-----------------------------------------

I think slowestTIP is more meaningful because we are launching speculative attempt for the
slowest task only. No ? Why do we call the tip selected(in this method) as "latestTIP" ?

> JobInProgress.findSpeculativeTask should use min() to find the candidate instead of sort()
> ------------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-1829
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1829
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>          Components: jobtracker
>    Affects Versions: 0.22.0
>            Reporter: Scott Chen
>            Assignee: Scott Chen
>             Fix For: 0.22.0
>
>         Attachments: MAPREDUCE-1829-20100610.txt, MAPREDUCE-1829.txt
>
>
> findSpeculativeTask needs only one candidate to speculate so it does not need to sort
the whole list. It may looks OK but someone can still submit big jobs with small slow task
thresholds. In this case, this sorting becomes expensive.

-- 
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