A. If you need to completely shut down your Hyper-V cluster, the key is to make sure that all guest OSs are shut down cleanly, which will enable them to start faster.
Shut down each guest OS from the Failover Cluster Manager console. Don't shut down within the guest OSs—if you do, the cluster service will think the virtual machine (VM) had a problem and restart it. Likewise, if you just shut down the host OS, the VMs may move to another node, or, if the host is the last in the cluster, then it may kill VMs if they don't shut down in a timely fashion.
Once all the VMs are stopped, you can (optionally) take the disk resources that host the VM resources offline.
Once these steps are done, you can shut down the node OSs. You don't have to follow a special order.
Once the cluster comes back online, you should try to start all nodes at the same time so you get better distribution and a shorter time for the VMs to come online. You should confirm the distribution once everything is running. If you're running a Windows Server 2008 R2 cluster, make sure that the Persistent Mode setting is enabled for all groups so that they return to their original locations.
Related Reading:
- Q. I'm performing a Live Migration or vMotion operation. During the migration, I lose a couple of ping packets. Is this normal?
- Q. I created a new virtual machine (VM) and I'm connecting it to the virtual hard disk of a previous VM. I still have the old VM's XML configuration file. Can I change the new VM so it keeps the old one's network configuration?
- Q. How do I set a CPU reservation for a Hyper-V virtual machine (VM)?
- Q. Can I set processor affinity with Hyper-V?
Check out hundreds more useful Q&As like this in John Savill's FAQ for Windows. Also, watch instructional videos made by John at ITTV.net.
