Mailing-List: contact users-help@activemq.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@activemq.apache.org
Received-SPF: softfail (nike.apache.org: transitioning domain of
 rburnham@console.com.au does not designate 216.139.250.139 as permitted
 sender)
Date: Thu, 6 Sep 2012 18:04:26 -0700 (PDT)
From: Ryan <rburnham@console.com.au>
To: users@activemq.apache.org
Message-ID: <1346979866752-4656130.post@n4.nabble.com>
Subject: Shared File System Master Slave locking issues with VM's
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi,

I'm currently in the process of setting up a VM image for a Queue server,
the idea is we can use it to scale load. They will share the same File Store
and we have it set up to use the Shared File System Master Slave. So one
should pick up as the master by locking the store and the other comes up as
a slave when it can't lock it.

The problem though is we had some very random results when testing it. We
were writing 2 messages and reading one, making sure the what we got back
was in sequence. Suddenly things went out of whack and when we looked into
it both the Master and Slave seemed to have queues running.

It wasn't until later that i started wondering if it was cause by me saving
the state of the VM. I saved the state of both the Master and Slave. Waited
a while and then started the Slave first. the slave took the lock and became
a Master. Then i resumed the old Master. It still seemed to think that it
was the Master. As far as the Old Master is aware the machine never stopped.  

So can anyone think of a way to help prevent this from happening in terms of
configuring ActiveMQ. We don't want to use Pure Master Slave because it has
no Auto Recovery


--
View this message in context: http://activemq.2283324.n4.nabble.com/Shared-File-System-Master-Slave-locking-issues-with-VM-s-tp4656130.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.