Mailing-List: contact yarn-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: yarn-issues@hadoop.apache.org
Date: Fri, 4 Jan 2013 21:18:12 +0000 (UTC)
From: "Jason Lowe (JIRA)" <jira@apache.org>
To: yarn-issues@hadoop.apache.org
Message-ID: <JIRA.12625800.1357176815843.83019.1357334292605@arcas>
In-Reply-To: <JIRA.12625800.1357176815843@arcas>
References: <JIRA.12625800.1357176815843@arcas>
Subject: [jira] [Commented] (YARN-308) Improve documentation about what
 "asks" means in AMRMProtocol
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/YARN-308?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13544263#comment-13544263 ] 

Jason Lowe commented on YARN-308:
---------------------------------

bq.  For pure absolute, we would only need to send the full list for the rare heartbeats where we want to make a change to an application's request. An empty list of asks would indicate no change, as it does currently.

Ah, I didn't think of that since it's not pure absolute then (since it special-cases empty).  But we could separate heartbeating from allocating to solve that or keep the special-case.  However the common case of small deltas to the ask also becomes cumbersome in absolute.  For example, the MR AM asks for a container on every node with a replica of an input split.  When it gets one of those containers, it modifies its ask to remove the other two containers.  If the AM gets one out of thousands of containers, it still needs to send thousands of asks to do the small update, leading to more load on the AM and RM for the common (at least in MR) case.

                
> Improve documentation about what "asks" means in AMRMProtocol
> -------------------------------------------------------------
>
>                 Key: YARN-308
>                 URL: https://issues.apache.org/jira/browse/YARN-308
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: api, resourcemanager
>    Affects Versions: 2.0.2-alpha
>            Reporter: Sandy Ryza
>            Assignee: Sandy Ryza
>             Fix For: 2.0.3-alpha
>
>
> It's unclear to me from reading the javadoc exactly what "asks" means when the AM sends a heartbeat to the RM.  Is the AM supposed to send a list of all resources that it is waiting for?  Or just inform the RM about new ones that it wants?

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira