Mailing-List: contact activemq-dev-help@geronimo.apache.org; run by ezmlm
Precedence: bulk
Reply-To: activemq-dev@geronimo.apache.org
Message-ID: <25780420.1155550823248.JavaMail.jira@brutus>
Date: Mon, 14 Aug 2006 03:20:23 -0700 (PDT)
From: "james strachan (JIRA)" <jira@apache.org>
To: activemq-dev@geronimo.apache.org
Subject: [jira] Assigned: (AMQ-878) JMX Exception: "Destination already
 created" when recovering gigantic queues from database
In-Reply-To: <31663514.1155370888562.JavaMail.jira@brutus>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

     [ https://issues.apache.org/activemq/browse/AMQ-878?page=all ]

james strachan reassigned AMQ-878:
----------------------------------

    Assignee: Hiram Chirino

> JMX Exception: "Destination already created" when recovering gigantic queues from database
> ------------------------------------------------------------------------------------------
>
>                 Key: AMQ-878
>                 URL: https://issues.apache.org/activemq/browse/AMQ-878
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Broker
>    Affects Versions: incubation
>         Environment: Embeded broker, Any platform (OS, JDK) - currentlly embeded in JBoss 4.0.4.GA
>            Reporter: Nikolai Penkov
>         Assigned To: Hiram Chirino
>         Attachments: async-recovery.patch
>
>
> I think this issue is described in Hiram's blog: http://blogbucket.blogspot.com/2006/05/scaling-to-gigantic-queues-and-topics.html.
> E.g. when broker is stopped with a lot of messages in queue. When started - recovery process is fired (currently on creation of a queue), and if another subscriber requests queue - it creates a new one, because previous process hasn't finished.
> I think the simpliest solution for that problem is to start recovery process in another thread with synchronization to sent process(sent process should be suspended till full recovery is done) and subscriber notification method (every sunscriber should be notified on recovery of message).
> I have attached a patch to activemq-core, which I have tested with my environment - JBoss 4.0.4.GA with Embeded Broker ActiveMQ trunk release, jdbc persistence.
> I am not a spec. in threads nor in ActiveMQ design, and I am sure that you will find a more elegant sollution to what I have provided.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://issues.apache.org/activemq/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira