I finally got my lab updated and the two vms started, so far so good. Stats data is present, the first time I've seen any of my hosts pull 100% CPU. Hosts are little HP T740 thin clients with AMD V1756B processors and 64GB of ram, in a 3 host pool.
Sorry I can't bring more detail, not really doing work on my XCP lab right now, and lots of shake up around tech. space at work.
I decided to burn my lab down because of some trouble and wanted to start fresh to isolate any problems I may have created in the past. Wiped the disks, built a new NFS share ready for work, installed the XCP-ng OS and configured networking for the management interface on the local machines, all normal stuff. Went back to my desk and logged into XO Lite on each machine, using console --> xsconsole I was able to get all the hosts into a pool, and add an ISO and NFS SR, all seemed to work.
Went to the New VM to create a simple Debian vm on which I intended to host XO-CE from sources. Typed the info in, chose the Debian image from my ISO SR, the rest of the config as normal. Uncheck the boot after creation box, and create.
Click on the VM to bring the console up, then click power on... UEFI loader comes up, and ... Can't find the boot disk. Tried another iso that I knew was capable of EFI boot, same.
Realized I was trying to start in EFI, made a new one in BIOS mode, this one worked but now my screen shots are messed up so I'll probably have to go through and delete stuff with a different XO and go back to grabbing images of the process to write up "for the next guy".
Other than this small issue, it's working and doing it all from XO Lite. Yes still a bunch of missing features, but I know you are working on this. I just wanted to try things this way on this fresh build and see what I could get done from local console and XO Lite.
HP T740 with 64g of ram, 256g sata for OS, optional nvme for faster storage, optional 10gbe networking through low profile card. Works great for my lab.
Ran vSphere8 on three of them, running Harvester on 3 of them now. XCP-ng and vSphere8 work great, Harvester is a bit more resource hungry. Only the bottom three have nvme and currently have Harvester 1.7.0 running while I try to find time to learn more about this system.
My production pool upgraded fine as far as I can tell, but not so good with the windows drivers, I'll make a thread on the drivers.
Nothing really to add, my 3 host Intel production pool updated just fine. The load balancer is always a little weird, but I'm sure it is calculated based on CPU and RAM assigned to each VM, where I split things up based on workload.
It's a small system, and the real workload is handled by 3 Windows VMs so I tend to split them up onto one of the three hosts.
I may get to my lab in the next couple of days, but it isn't doing work so testing is kind of pointless right now. The only thing "doing work" is a VM with XO from sources.
BTW, the ram increase made zero difference in the VDI migration times.
I may still change the ram for each host in production, but I only have like 12 VMs right now and probably not many more to create in my little system. Might be useful if I ever get down to a single host because of a catastrophe. Currently running what it chose for default at 7.41 MiB, roughly doubled that in my lab to see what might happen.
Just got my VMware lab started so all this other stuff is going to back burner for a while. I will say I like the basic ESXi interface that they give you, I can see why smaller systems may not bother buying VCenter. I hope XO-lite gives us the same amount of function when it gets done (or more).
I was looking for a way to mark this solved, can't find it.
I haven't moved things, but after migrating my big lab to my mini-lab, I'm confident that the warm migration is the way to go. It was fast and seamless as long as you have the right network adapters set up. I had to fool with one of my networks to make a VM function, but that was certainly something I overlooked while setting up the mini-lab. A little testing before moving the VMs should make this go easily if using the old servers is the option for this project.
Thanks. I'm still going to wait until next week, my schedule just doesn't give me the time right now. But nice to know that when I do update, it will be fixed and my laziness prevented a problem for me 
I think I'll get my pool updated today, only been waiting a week at this point.
@coolsport00 You actually referenced my post, so I referenced yours as backup to the condition still being present. I think we just need to wait for them to finish making XO-lite what they seem to intend it to be, I think it's going to be good and I think they are doing a great job so far.
Also, I have a post somewhere about the UEFI not working from XO-lite, maybe they can be merged?
That's an interesting way of doing this, I'll have to remember it. I think the Fuel Ignition tool has a vhd output, it also produces both ignition and combustion parts for you. Worked pretty good on one test machine, but I need to go deeper into some of the other choices you can type in.
This is mostly a thought exercise for discussion, and maybe you've already thought about all this and made choices based on performance or other metrics.
Since version 9 is being designed, and it's being designed around Alma 10 as a base, would it make sense to revisit the HCI options?
Everyone seems to scream for Rook/Ceph. I've never used it so not sure.
Some people want Kubernetes or some container system installed.
What about setting up Kubernetes (k3s or rke2) and using it to provide Longhorn as storage?
Do I know what this would take to get going? No, not yet. I've been working with Harvester which uses Longhorn V1 or experimentally Longhorn v2. I'm also finding out that you really need to have Kubernetes running and have Rancher running to manage more aspects of Harvester without going to yaml or Helm scripts. So I'm also diving down the must learn a bit of Kubernetes so I can get Rancher so I can use Harvester the way other platforms work. It's a rabbit hole.
Downsides of Longhorn or maybe just downsides of Harvester, can not use NFS "out of the box". You need to install the CSI driver for NFS, and then it is only for backups, you can also install S3 and do the same. Longhorn is also fairly slow, even going across a 25gbps connection to an nvme. The local performance of this nvme has been tested to 3GBps with ESXi, so not sure why I can't get faster, but not the point of this discussion.
Just though I would bring up another choice that I've never seen discussed, and throw it to the winds for consideration going forward. V9 is so young that major changes could be accomplished if they made functional sense. I know you spent years working with Linbit to get what we have now, and I've not tried this either so hard to say anything good or bad about it.
Replying to an old thread because this one popped up in a search.
There is an open source and locally running ignition/combustion generator that I've used on bare metal. To use it with XCP-ng, we would need the ability to have a second "cd rom" drive during the install phase, one for the OS installer, one for the combustion/ignition ISO generated from the tool.
https://opensuse.github.io/fuel-ignition/
They say it only runs locally, no info is send out to the web. I didn't monitor my connection on the one time I used it, and probably won't the next time (coming soon to a bare metal cluster in my lab). For the record, I'm using the LEAP version of Micro for my tasks.
The ronivay script requires you to select an option (#2 to update).
I look at things this way, it's good to have more people working on scripts like this.
I haven't tried this yet, but liking the menu you just showed!
How many of these are critical? I haven't even had time to apply the last round of patches to either lab or production. 
In theory Harvester supports nested virtualization, but I haven't tried it yet. Might be worth setting up a single host to try it. I think Proxmox also supports it, again might be worth a single host if that would handle the load.
Do you have compression turned on for the backups?
I have a couple vms that I don't compress because of the old error, I didn't update yet or Red enable compression to see what that status is.
Normal for the pool master to take a bit longer but never seen 15 minutes before.
The wait period will be nice, after an RPU I'm watched VMs bounce back and forth trying to put themselves on another host. It's been a few updates since I noticed that, but a cooldown seems like a good idea, especially one that can be adjusted as you've made this one.