We are testing a VDI deployment at the early stages and one sticking point at the moment is how are we supposed to safely drain the RDVH host server to reboot for windows updates every month?
We have 2 RDVH servers in separate data-centers for resilience, there is no shared storage and trying to 'move' a VM fails (although this would not be ideal anyway).
I have looked for the equivalent of 'Allow new connections' for RD session hosts or a load balancing option where we could put all new connection requests onto one server whilst the other one drains but I cannot find a solution to this.
We operate 24x7x365 so we have no maintenance windows for downtime on this, nor should we require one with 2 Geo-separated servers.
One possibility would be to use powershell and a loop to:
- Save the state of unused VM's
- Delete the VM's when in a saved or Off state in Hyper-V
- Reboot the server when all VM's are gone
- Create new VM's when the reboot(s) are complete
I would expect there to be a simpler way, but I can't find one. Please help!