I am looking into the possibility of using Hibernate 3 to develop a database persistence layer to be deployed on a clustered Resin Server(3.X).

Most of the documentation I have found on the net seems to deal with using Hibernate in a standard j2se situation so when it comes to deploying it in a j2ee, clustered environment to take advantage of the containers connection pools and connection management I am a little lost.

When deployed in a cluster, does that imply that for every node in the cluster I will have a unique instance of the SessionFactory - surely that can't be good if it is the case. How can it be avoided???

Any ideas and suggestions with this regard would be greatly apreciated.