Trying to wrap my head around the best way to test a DR failover when there is a mixture of physical servers and virtual machines that are part of the failover. Here is the scenario, app / web servers all run in VM's that will be protected via SRM, they run on a mix of 3 VLANs then there are 4 database servers that are running on a database VLAN. The plan is to use native replication within the database programs for data replication, and keep 2 virtual domain controllers in a separate site but part of the same domain.
If I put all traffic on some isolated network segment so everything in the failover site can talk to each other but not talk to the production network, I would think that would work but I have concerns about how my two AD servers will respond that are part of the protected site, should I just put them as a protected VM so the test recovery can just throwaway any changes that are made within AD at the time of the test?
Guess I'm just looking for some real-world this is how we did it in a mixed (physical / virtual) environment.
