today another story I see most of the days when I do Healthcheck on customer site.
One of the first things I found was a new VMM installation on top of a new Microsoft Hyper-V Cluster. The issue was, that the cluster was installed first and every thing was running on Hyper-V converged network and standard Hyper-V Switches. There was no VMM network configured or the host made compliant within VMM.
Why is that so bad? It’s bad because VMM uses it’s own kind of switches (logical switches) and needs some additional configurations to manage the hosts in the best possible way.
When I ask guy’s why they do it in that way, I normally get the answer “how should I configure the host when no VMM is in place before he install the cluster?”.
So now my answer and how you can do it in the right way.
- Install a Hyper-V Host as Standalone host
- Configure and install the VMs for VMM and SQL Server (if needed) on the standalone host
- Performe the ful VMM configuration
- Install the other Hyper-V Hosts and roll out the VMM configuration to that hosts
- Cluster the Hyper-V Hosts
- Migrate your SQL DB and VMM with shared nothing livemigration to the new Hyper-V Cluster
- Reconfigure the standalone Hyper-V Host with VMM and add it to the cluster
- Run the cluster validation again
That’s all and it will cost you the same amount of time.