Mailing-List: contact users-help@activemq.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@activemq.apache.org
Received-SPF: pass (athena.apache.org: domain of lists@nabble.com designates
 216.139.236.158 as permitted sender)
Message-ID: <18699994.post@talk.nabble.com>
Date: Mon, 28 Jul 2008 14:36:34 -0700 (PDT)
From: skomarla <skomarla@yahoo.com>
To: users@activemq.apache.org
Subject: Re: Master/Slave JDBC persistence using embedded broker in JBoss
In-Reply-To: <18699988.post@talk.nabble.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
References: <18698970.post@talk.nabble.com> <18699988.post@talk.nabble.com>


I did see that but I guess I was hoping for this to be a little more elegant. 
I feel that having to define all the failover URLs defeats the purpose of
in-vm protocol in the case of an embedded master slave.  

It would be have been nice if the in vm broker is started in a "slave" mode,
where it would forward the messages to the master.  Perhaps it was my
misunderstanding of ActiveMQ, but I thought broker discovery would have
helped achive this.  I can't use networked brokers because I want all the
nodes to share the same database.


Hans Bausewein wrote:
> 
> First of all, I'm not sure whether I'm the right person to answer your
> questions.
> There may be more uses of ActiveMQ than I know of.
> 
> I've deployed 5.1.0 and the development version in a clustered JBoss
> environment with pure master/slave and as a single broker (because of some
> issues). 
> 
> 
> skomarla wrote:
>> 
>> Hello,
>> 
>> I'm having trouble getting a master slave setup working using an embedded
>> broker in JBoss with JDBC persistence. I've done some searching to see if
>> there is a workaround, but the only thing i've seen seems to be to either
>> run master/slave separately, which is what I'm trying to avoid.
>> 
>> The overview of the architecture is this.  
>> 1) There will be multiple JBoss instances each hosting a spring based
>> application, which contain JMS listeners (using the spring's
>> org.springframework.jms.listener.DefaultMessageListenerContainer).  
>> 2) In order to keep the deployment as simple as possible, we have decided
>> to use the embedded activemq broker and have each node use the in vm
>> transport protocol.
>> 3) I'm using activemq 5.1.0, and so I don't have a problem with the slave
>> node completing its startup sequence.
>> 
>> My problem occurs when my spring based application deploys.  
>> - The master node completes its start up sequence and the application
>> deploys.  
>> - The slave node complete's startup sequence without blocking, but when
>> the application deploys, the DefaultMessageListenerContainer repeatedly
>> tries to connect to the broker.  This fails with the below exception.
>> (seemingly because the slave has not started its transport connectors
>> which it seems to only do so when it becomes a master)
>> 
> 
> The documentation is quite clear:
> 
>   http://activemq.apache.org/jdbc-master-slave.html
> 
> "Only the master broker starts up its transport connectors and so the
> clients can only connect to the master."
> 
> If you want to know why, one of the designers/developers probably better
> answers that.
> 
> I guess it's a lot easier to accept messages on one entry point only.
> 
> Hans
> 
> 

-- 
View this message in context: http://www.nabble.com/Master-Slave-JDBC-persistence-using-embedded-broker-in-JBoss-tp18698970p18699994.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.