The release we're working on (1.2) only deals with clustering in a LAN enviroment. We rely on the fact that either we have access to a shared database, or we can use hardware multicast for in-memory replication. Either prerequisite is only satisfied in LAN. We have plans to extend clustering for WAN in 2.0, but they are not very well defined at this stage. In our design discussions, we envisioned a scenario similar to what you've described, but we won't provide support for that in 1.2. Regardless of when it will be implemented, it will use JGroups underneath. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3973376#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...