I'd like to clarify that I have a hardware LB in front of two separate Apache + mod_jk (configured to load balance) nodes. Each Apache + mod_jk node lb's to the same two JBoss AS servers. If the HW LB cluster in front of the Apache nodes decides to send a request to a node other than the originating Apache + mod_jk sticky session initiating node (we'll call this ws node 1) during an active session then my assumption is the other Apache + mod_jk node (ws node 2) will have no knowledge or ability to handle the other Apache + mod_jk lb nodes activities/ sticky session? Thus my reason for skipping the sticky sessions. Follow me? The one thing they do have going for them is that they are both configured identically mod_jk wise. The workers are defined the same so therefore the jvmRoute entries theoretically could be identified by ws node 2 if ws node 1 wasn't available. I guess this brings me to another question. Does mod_jk keep track of sticky sessions it's initiated locally? How is that information stored/shared? If there are two separate web servers with mod_jk identically configured to LB w/ sticky_session=true and sharing the same workers, could one entire web server fail and the other take over seamlessly? How would JBossAS handle this. How would mod_jk handle this? View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4242491#... Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&a...