The principle reason to distribute your implementation is local survivability. You can distribute your entire network and place application servers at all location or just those segments where messaging service should not be interrupted. By placing applications servers at remote sites, you can ensure continuation of messaging service in the event of a network outage between the remote site and the main location. If the storage server at the main site becomes unreachable to the local server, users at the remote site experience virtually no difference in service and can retain user messages of the previous 72 hours. Another benefit of the distributed multiple site configuration is that you have less voice traffic on the WAN. A possible trade off is that you might not always be able to maximize server capacity.
