We are currently running DFS between several of our remote sites. We are basically using DFS to replicate the remote sites data back to our main office to be backed up. Right now all of our DFS replications are connected in a 1 to 1 topology. The replication group only consits of two servers, the remote file server and our file server in the main office.
We are preparing to deploy a new file server to another remote site which is in close proximity to one of our other remote sites. It is forseeable that users could on any given day be working at either office.
Right now the new file server that is going to be built is in the same office as an existing server that we are going to replicate data to. What we would like to do is set a replication group between these two servers and let them utilize the gigabit lan speed to perform their initial replication.
Once initial replication is complete we would move one of the servers to the new site which is connected via T1. There should only be minimal traffic since the initial replication was done already. Once the two servers are in the different site we would like to add a third member server into the replication group in our main office to be able to backup the data.
Hopefully my terrible drawing helps you visualize what Im trying to do
So now my questions
1. We are planning on adding the replication members in two steps to take advantage of the ability to use FULL bandwidth when in the same LAN. Once the servers are seperated we will use a throttled schedule. Do you see any problems with this?
2. Can you start a DFS replication group and then add a third member after its initial replication?
3. Any other information I should be aware of?