[cisco-voip] Off site subscriber for disaster recovery

[Chris Parker]

Hello,

I am working on a design for disaster recovery for a Call manager and 
Unity connection cluster. The idea is to have one sub located off site 
in a remote data center for both the Call Manager and Unity clusters.

Neither sub would ever process any calls unless the production site was 
down. I am wondering how much bandwidth I am going to need. Since the 
Call Manager sub isn't processing calls, it should only need to receive 
data from the pub when there is a configuration change in the database? 
So I'm thinking the traffic will be pretty low.

The UC sub may require a lot of bandwidth if messages need to be 
replicated. Is this done in real time or are the messages just copied 
over using ftp etc.? How much bandwidth does the server really need? I 
suppose I can look at how many messages we get in a day and calculate 
the file sizes add them up etc. But I don't know what kind of time frame 
the messages need to be copied for UC to be happy. Since its for DR I 
mainly care about the configuration being correct on the UC sub, and not 
so much the mailbox contents. Is it possible the have a UC sub that only 
syncs configuration and not messages?

I know the SRND's give numbers for both these scenarios, but they seem 
really high. Anyone have experience doing this?

Chris