To elaborate, it feels wrong that if you set up a bridge to send messages "somewhere", to some remote JMS provider, that you have to know details about the way that remote provider is configured. It feels to me like you shouldn't have to worry about which ServerPeerIDs that remote provider is using internally. As far as I can see, it should be just a "black box" remote JMS system. Ideally, you shouldn't even have to think about whether it's JBM or some other JMS implementation. That's why I think the right solution would be not to tell people "whenever you have two JBM instances talking to each other, whether it be clustering or bridging, make sure every node involved has a unique ID". I think JBM code should distinguish between the distributed case and the bridged case when it comes to clusters and node IDs. But again, pointing out the fact that this is required would obviously already be a step forward from the current situation. I guess it would also be good to mention somewhere in the documentation that JBM bridging uses a "special case" when it talks to another JBM instance. (At least that's how I understand the situation.) That's just my $0.02 though... Cheers, Julian View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4087514#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...