"Clustering with JBoss 3.0" article questions

[joao.clemente]

Hi.
I've read the article posted on O'Reilly and I got a couple questions.

I can't understand the reasoning under the sentence "since JBoss recommends running your servlet engine (Tomcat or Jetty) within the same VM for performance reasons, you really don't need EJB fail-over or replication either"
Can someone explain this better?
I understood the sentence that appeared before: If you can live with forcing a user to log in again after a server failure (you can afford lose the http session info)

Maybe Bill Burke or Sacha can explain what they meant a little further?

[slaboure]

What do you want to failover? your state? ok, then, in a web application, the state is generally in the httpsession (or in a SFSB).

Consequently, if your http-session is replicated and you have an homogeneous deployement of your app in all nodes of the cluster, you don't need anything else: the servlet will always access the local EJB and never fail to another node (no need for this).

Is that clearer?

cheers,



Sacha