Mailing-List: contact mapreduce-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: mapreduce-issues@hadoop.apache.org
Message-ID: <7485140.181621288643607639.JavaMail.jira@thor>
Date: Mon, 1 Nov 2010 16:33:27 -0400 (EDT)
From: "Allen Wittenauer (JIRA)" <jira@apache.org>
To: mapreduce-issues@hadoop.apache.org
Subject: [jira] Commented: (MAPREDUCE-2170) Send out last-minute load
 averages in TaskTrackerStatus
In-Reply-To: <17678413.180651288641984842.JavaMail.jira@thor>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/MAPREDUCE-2170?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12927145#action_12927145 ] 

Allen Wittenauer commented on MAPREDUCE-2170:
---------------------------------------------

That's sort of my point:  why not promote this to be a Hadoop feature rather than something tied to a specific OS if we can support it in a generic way?

One of the bad things about load avg is that it doesn't give you a why.  In other words, a node may be totally fine running more cpu-bound tasks if the box is actually being hammered IO-wise.  Since both conditions show as load....  But this would require a lot more smarts in the system and of the user.  So while it makes me slightly uncomfortable to use this value, load should be 'good enough' given the current state of the system.

> Send out last-minute load averages in TaskTrackerStatus
> -------------------------------------------------------
>
>                 Key: MAPREDUCE-2170
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-2170
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>          Components: jobtracker
>    Affects Versions: 0.22.0
>         Environment: GNU/Linux
>            Reporter: Harsh J Chouraria
>            Assignee: Harsh J Chouraria
>            Priority: Minor
>             Fix For: 0.22.0
>
>   Original Estimate: 0.33h
>  Remaining Estimate: 0.33h
>
> Load averages could be useful in scheduling. This patch looks to extend the existing Linux resource plugin (via /proc/loadavg file) to allow transmitting load averages of the last one minute via the TaskTrackerStatus.
> Patch is up for review, with test cases added, at: https://reviews.apache.org/r/20/

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