[jboss-user] [Clustering/JBoss] - JNDI vs. HA-JNDI lookups

Hi, I've been using JBoss clustering for a while now and it's working great. Thanks! Now I've come across a matter from which I thought I knew how it worked for sure, but somehow it doesnt. It's about the behaviour of (HA-) JNDI lookups from a remote client. My assumption was the following: you have two ports on each node, one exposing the local JNDI tree (usually ports 1099, 1199, ...) and the other one exposing the HA-JNDI tree (usually ports 1100, 1200, ...). Now when a client connects to the server using the port for the local tree (1099), the request should be handled on the node it called, and on that node only. When the client connects to the server using the port for the HA tree (1100), the request will be handled by any node in the cluster. Or, to be more precise, the smart proxy, which the client got with the first call, will distribute the calls across the nodes. I'm pretty sure that's the behaviour I saw the last time I checked. But what I'm seeing now (we changed to JBoss 4.0.4 in the meantime) is that also the calls to the local port (1099 etc) are distributed across the nodes. However, for testing purposes, I need to call a certain node, and not any node in the cluster. Is my understanding wrong (if so, what's the right way of doing it)? Or was there a change with JBoss 4? Or is there even a bug in JBoss 4? Any enlightenment is appreciated. :-) Martin View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4024269#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...