Hi samdoyle:
It can be a bit of a transition from single-instance based development to clustering.
I suggest you spend a little quality time going over the HA Admin Guide reference I mentioned in an earlier post. It covers many of the issues that you are starting to
confront, including setting up load balancing, etc. It's worth downloading and having
at your side as you work through. Here's the reference again.
One other important thing that people sometimes forget about when transitioning from single instance development to a cluster is that all of the objects you put in http session must now be serializable. You never have to worry about that in single instance development but it will hit you when you do clustering because the object graph under the http session is serialized, replicated and upon subsequent load de-serialized when needed. Anything that can't be properly serialized or de-serialized will suffer.
https://glassfish.dev.java.net/nonav/javaee5/docs/SJSASEEHAAG.pdf
[Message sent by forum member 'lwhite' (lwhite)]
http://forums.java.net/jive/thread.jspa?messageID=248134