I have a question on how SQL Server 2012 handles data flow with multiple replicas in an availability group.
Here is the scenario:
A primary and a secondary AG replica is the primary site. Also two replicas go to the remote DR site. The reason for two replicas going to remote site is in the event of a DR failover one replica would be the intended primary and the other would be for reporting.
The question is how does SQL Server and/or WFCS handle the data flow for the primary to the two remote replicas. Specifically is the data flow over the WAN link duplicated from the primary going to both remote replicas? Or is it somehow smart enough to transfer the data over the remote WAN only once and split it at the remote end once it gets there?