Thank you very much for your detailed reply. I will check out those articles you linked to. As to the later statements that weren't to clear, yes, we are using NetApp and their SRA, and we do run SnapMirror jobs daily for our database server, which does use RDMs. When running a test of our Recovery plan, I see the datastore which contains the LUNs, and I see the drives mapped on the VM, but SnapDrive is not presenting the drives, I'm not sure how they are attaching. I try to connect the drives within SnapDrive, but it can't find the storage array...probably because it is in a network bubble, and yet, the drives are there and the letters are correct when comparing to the production DB. I'm thinking the SRA is actually handling that (?)
I can tell the VMs are old because every time I run the test it asks me to apply the same Microsoft Updates...and I understand this is read only, so changes I make will not persist across tests after running a cleanup. I think I need to replicate the data, which I have not been to save time. That will likely answer that question.
I was also told by other sources to replicate an AD server, as right now when I console in, it is using cached accounts. In trying to emulate a better test environment, I was going to use IP customization and assign the Recovery VMs to the local subnet in our Recovery site, but I think the vSwitch still isolates them regardless of their IP address. I assume I still won't be able to talk to any production servers that are on the same vCenter server or get to the internet. If there were an option to use a test network that isn't isolated, I could do that since the IPs I've customized are unique, no clashing would occur.
Again, I sincerely appreciate your response, it did clarify several things for me.