hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Wangda Tan (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (YARN-4524) Cleanup AppSchedulingInfo
Date Tue, 29 Dec 2015 22:38:49 GMT

    [ https://issues.apache.org/jira/browse/YARN-4524?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15074375#comment-15074375
] 

Wangda Tan commented on YARN-4524:
----------------------------------

Thanks [~kasha] working on this, this is very useful for our future works on scheduler, some
comments:
- By design, only increase request will be cached in AppSchedulingInfo (all decrease requests
will be handled immediately when AM make allocation call just like release containers). So
maybe changes from {{-increase-}} to {{-resize-}} may not be accurate.
- In {{move}} call, not sure why following lines are removed:
{code}
738	    oldMetrics.moveAppFrom(this);		
739	    newMetrics.moveAppTo(this);		
740	    activeUsersManager.deactivateApplication(user, applicationId);		
741	    activeUsersManager = newQueue.getActiveUsersManager();		
742	    activeUsersManager.activateApplication(user, applicationId);		
743	    this.queue = newQueue;		
744	    this.queueName = newQueue.getQueueName();
{code}

> Cleanup AppSchedulingInfo
> -------------------------
>
>                 Key: YARN-4524
>                 URL: https://issues.apache.org/jira/browse/YARN-4524
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: scheduler
>    Affects Versions: 2.8.0
>            Reporter: Karthik Kambatla
>            Assignee: Karthik Kambatla
>         Attachments: yarn-4524-1.patch
>
>
> The AppSchedulingInfo class has become very hard to grok with some pretty long methods.
It needs some cleaning up. 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message