I'm testing out Neil Peterson's method of patching Clusters with Orchestrator and SCCM 2012. In his example, he has two nodes. He drains one node and patches it. Then he does that on the other node. All the VMs are configured with a preferred owner setting so after all is said and done, all the VMs return home and all is well. It really is pretty slick.
I'm wondering the best strategy for an 8 node Cluster. His example seems like it would work perfectly for that too. My big concern is that when the VMs get moved they won't all get stacked up on one node and overload it.
Orange County District Attorney