Ok seeking some clarity on why you would use a ClusteredCacheLoader in my case. I have seen the thread http://www.jboss.org/index.html?module=bb&op=viewtopic&t=118598 but in terms of my needs ;). Ok running my cache in invalidation mode is a no go for me, i dont want data evicted if i update a data item on one cache node. So i cant run in invalidation mode. replication mode without buddy replication. Doesnt make sense to use it here because all the data is replicated across all the nodes. replication with buddy replication: This is where i am not so clear, im not so clear on how buddy replication works, i have set it up but dont know the reasoning behind why it does what it does. Ok i would have thought that you would share data with your buddy and the data would be in the same tree, not in a seperate branch i.e under /_BUDDY_BACKUP_ 0 branch. Maybe im missing some configuration setting to merge the 2 buddies data into the same tree?? If i dont have buddy replication then all the replication goes into the same tree, thats how i would have thought buddy replication would work except that you are just sharing with you and your buddy has and not with the whole cluster. Can someone just clear up why 2 different trees and not the same as in replication without buddy replication?? p.s i using jmx console to see whats in my cache View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4108808#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...