activemq-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From jcarreira <jcarre...@gmail.com>
Subject Getting started with ActiveMQ
Date Wed, 26 Jul 2006 16:43:58 GMT

I'm getting started setting up ActiveMQ in a Spring 2.0 based app, so I'm
configuring it in Spring's XML files (any chance of a Spring 2 configuration
extension XSD for ActiveMQ?).

What I'm trying to do is this:

The app will be deployed in a cluster, backed by Tangosol Coherence caching,
and each node will be independent. I want to be able to drop in new boxes to
join the cluster automatically. JMS in this case is being used to integrate
with another product which will be deployed separately on JBoss with
JBossMQ. They're building in their side of the JMS integration now. I want
to own the Queues on my side, and also have them own the Queues on their
side, and do JMS-to-JMS bridging. This will allow for either side to go down
(although realistically, I expect if anyone goes down it will be the single
JBoss box rather than the cluster on my side) and still be able to Queue
messages to be handled when the other side comes back up.

So I've been looking at the ActiveMQ docs, and I'm a little confused.

- What's the difference between peer and discovery? I suppose discovery can
be used for more than setting up a peer topology. Why doesn't the peer
configuration use discovery as one of the mechanisms for finding peers? I
definitely don't want to use multicast for all communication between cluster
nodes.

- Since peer doesn't use discovery, are there good examples of setting up
discovery to automatically manage cluster nodes where each uses an embedded
broker? This seems like exactly what peer does, but without discovery
support...

- How does one configure the broker and connection factory in Spring? I have
the config for the Broker, but there's no relationship / constuctor to
relate the two. Do I just need to set them up with the same transport
connector URIs?

- I'm confused by the docs about using a SAN for journal files and the
master-slave docs... Can I set up the nodes to all use a shared drive from a
SAN for the journal and all use the same journal files, or do they each need
their own journal files? I plan to back this with a DB (SQL Server in this
case) but in the interim if a machine goes down, I'd like to not have the
messages in the journal file lying dormant. 

Ok, I guess that's enough to start... 

Jason
-- 
View this message in context: http://www.nabble.com/Getting-started-with-ActiveMQ-tf2004874.html#a5506781
Sent from the ActiveMQ - User forum at Nabble.com.


Mime
View raw message