@jedimarcus perfect, thank you for the confirmation!

Hello,
So our pool master we down and isn't coming back up. (I have to reinstall XCP-ng onto the server now that it's fixed).

There were no VMs running on the master when it went down or any VMs on its local storage. (we saw the issue coming)

I have run the following commands on a slave and connected to it in XOA and it shows it as the pool master now:

xe pool-emergency-transition-to-master
xe-toolstack-restart
xe pool-designate-new-master host-uuid=<Slave UUID>
xe-toolstack-restart

What are my steps to get the other back onto the pool? Should I forget the old master from the pool and have it rejoin the pool after installing a fresh copy of XCP-ng or something else?

Okay, this makes sense, thank you for the clarification. The reason I was asking I've been having crazy issues with xcp-ng on my server with VMs crashing all the time on boot etc and have been trying to troubleshoot and just wanted to look into this as I've done every other thing I can possibly think of. thanks again!

It's Windows 10 Pro, I believe I've also seen the issue with Server 2019. I don't suspect it's a licensing thing due to the fact that I could just move the topologies to 2sockets / Xcores per socket

Is it normal for Windows VMs to only ever show 2 sockets in the VM itself? For example, if I give the VM 24 cores in XCP-ng and assign the topology 12 sockets/2 cores per socket the VM only shows 2 sockets, 4 cores. If I change it to 4 sockets/6 cores per socket it will show 2 sockets/12cores per socket.

Host info:
Dell Poweredge R730XD
2x Intel E5-2680v3 2.50Ghz
128GB RAM
12 2TB SAS Drives in RAID10 done through PERC H710 Mini