Mailing-List: contact yarn-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: yarn-issues@hadoop.apache.org
Date: Wed, 9 Dec 2015 22:32:11 +0000 (UTC)
From: "Wangda Tan (JIRA)" <jira@apache.org>
To: yarn-issues@hadoop.apache.org
Message-ID: <JIRA.12918093.1449169796000.318028.1449700331189@Atlassian.JIRA>
In-Reply-To: <JIRA.12918093.1449169796000@Atlassian.JIRA>
References: <JIRA.12918093.1449169796000@Atlassian.JIRA>
 <JIRA.12918093.1449169796741@arcas>
Subject: [jira] [Commented] (YARN-4416) Deadlock due to synchronised get
 Methods in AbstractCSQueue
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/YARN-4416?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15049493#comment-15049493 ] 

Wangda Tan commented on YARN-4416:
----------------------------------

Thanks for sharing your thoughts, [~Naganarasimha]/[~sunilg].

Looked at code, I think we need to be very careful with the locking changes of OrderingPolicy. Since it will likely cause CME in the future.

I would prefer to split the JIRA into two parts:
- Remove redundant locks, such as getAbsoluteCapacity.
- Improve locks of OrderingPolicy. Even if it closely related to LeafQueue, but I think we should try best to decouple it from LeafQueue to better API design. Potentially we need to rethink API of OrderingPolicy.

I suggest to convert both JIRAs to sub jiras of YARN-3091.

> Deadlock due to synchronised get Methods in AbstractCSQueue
> -----------------------------------------------------------
>
>                 Key: YARN-4416
>                 URL: https://issues.apache.org/jira/browse/YARN-4416
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: capacity scheduler, resourcemanager
>    Affects Versions: 2.7.1
>            Reporter: Naganarasimha G R
>            Assignee: Naganarasimha G R
>            Priority: Minor
>         Attachments: YARN-4416.v1.001.patch, YARN-4416.v1.002.patch, deadlock.log
>
>
> While debugging in eclipse came across a scenario where in i had to get to know the name of the queue but every time i tried to see the queue it was getting hung. On seeing the stack realized there was a deadlock but on analysis found out that it was only due to *queue.toString()* during debugging as {{AbstractCSQueue.getAbsoluteUsedCapacity}} was synchronized.
> Still i feel {{AbstractCSQueue}}'s getter methods need not be synchronized and better be handled through read and write locks.


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