Question for those that experience with Hyper-V / SCVMM in larger enterprise environments… I was tasked with architecting a new Hyper-V / SCVMM environment. We are running System Center 2012 R2, have 140 hosts, 650 guests spread around the world. In our main data center we have a production cluster, development cluster, DMZ cluster, etc. We currently have a single logical switch with several different port profiles across all these clusters, we do VLAN tagging on the vNIC’s. Most all hosts have 2 10GB NIC’s on them that are teamed active/standby.
I’ve read several articles about this and most say to run only one logical switch, but these articles aren’t written for larger environments in mind. Is there any benefit to creating a separate logical switch for each cluster? That way Production has logical switch, Development would have its own and down the line. Is there a point where a single logical switch would be taking on too much in the way of load/traffic?
Thanks!