@Bastien-Nollet Thanks as always @florent

Hello everyone, I absolutely understand the point regarding datacenters. I am using xcp-ng for years in smaller - non datacenter - environments and its great. Therefore I added the nut packages (8.2 and 8.3) and it works like a charm. It's installation is well explained in: https://xcp-ng.org/forum/topic/4300/performing-automated-shutdown-during-a-power-failure-using-a-usb-ups-with-nut-xcp-ng-8-2/13?_=1742229073030 However, I agree with @Kajetan321 that it would be great if the nut package could be included in the standard "updated" packages repo since its added quite some benefits (imho) for "smaller" IT environments. However, I do not know how much effort it takes to maintain that.

@Andrw0830 I am using paid version of ESXi / vSphere. The free version of ESXi has so many limitations that I consider it pretty much unusable / only usable for very limited basic tests. It would not suprise me if your issue is due to the license limitations within ESXi.

@Greg_E thanks for the clarification.

@CodeMercenary hello and thanks for this idea! We will discuss about it with @clemencebx and see what we can propose on new XO-6 design and workflow.

@lsouai-vates Great! Thanks for addressing this

It's not the first approach, you have KubeVirt for example.

Sometimes it when asking questions as well as hallucinating ChatGPT, can respond with results based on old versions (not updated for an up to date code base). So if any past in development or experimental code branch, had its code and was mistakenly released. Then the mistake was found and fixed, it can have found its way into the training data set. Can also be an example of training data set poisoning, forcing it to give misleading or mistaken responses, due to it hallucinating as a result.

@Andrew said in XOA suddenly "empty" during rolling update: @halvor Yes, kind of... The pool master must be updated and rebooted first. If you move the master to a different host then you are no longer rebooting the pool master first. XOA is just a very smart client and needs the pool master to provide all of the pool information. Secondary machines will just redirect to the master, so if you try to connect XOA to one as a backup it won't work. You just need to wait for the master to reboot and return to a functioning state and wait for XOA to automatically reconnect. The everything will start happening again. If your master host is destroyed (ie. software corruption, hardware failure) then an existing host can be forced to take over the pool as master, but this is not a good thing to do unless the actual pool master is unrecoverable. Ok, great, thank you for clearing this up!

No luck, I've tried building from source, using the script, etc. Following along with Google Gemini I even managed to get walked into the process of damaging the Xapi to the point of losing access to the server from Xen Orchestra / XCP-ng Center (unfortunately for me it was the Pool Master). Thankfully I had backups on my NAS and experienced minimal loses through the process of having to reinstall the operating system. At this point, I'm guessing my pool that dates back to version 8.1 is screwed up in some way beyond human or AI understanding. I have 2 Servers & 3 NAS units (2 TrueNas Scale) + 1 Synology that all work with the old version of Xen Orchestra but I can't get it to work with anything else. Manually putting the same info in or importing the configuration. One of the servers was rebuilt today and added back to the pool. Is it possible that reinstalling the OS on the other server could fix everything? I'm confused because the backups aren't failing from the old Xen Orchestra. Old: [image: 1756784257357-19e6f3da-3b22-4f7d-a47e-66d8216b3f8b-image.png] New: [image: 1756784292273-263493de-a5cb-4610-b7ba-ba173fe7b055-image.png] Any help and / or ideas would be greatly appreciated. Thanks!

Thanks everyone for their help. I believe I have a way forward now.

That's weird, but OK. Maybe it's because even when you update, the old/existing RRDs data won't be "replaced".

@olivierlambert Great to know it's not something I was doing wrong! Hopefully these other cards I have ordered will afford me more luck! Cheers

so it seems like when i make the igpu pass through, i see the starting lines of xcp-ng but it never gets to the console screen. which is what i expect. When I add the pcie devices (igpu/hdmi audio) to a vm and start it up, I get and update to the screen and it shows a red number counting which is part of the console. So it seems that when the VM re-initializes the igpu, dom/xcp retakes it over. I confirmed that by using the keyboard to move around and it reshowing the console and filling the missing details as i move around. So why is the dom/xcp retaking the video device over again when it should be blocked from accessing it? [image: 1756568880819-img_3348.jpeg]

@planedrop While you are connected to the XOA VM, be sure to verify that the required changes have been performed on /etc/systemd/system/xo-server.service when the VM memory was increased. Details can be found in this troubleshooting KB article -- https://help.vates.tech/kb/en-us/31/48

@Greg_E Ah, ok..yeah...guess I did Yep, agree. Really needs more functions...like networking (and elsewhere) ha

I have not had time to work on this, and my lab changed significantly, had to spend a bunch of money to move it to a new rack and that rack has been disconnected for months so no real work in XCP or vSphere and learning to migrate VMs from vSphere to XCP which was one of my big intentions. But a recent post confirms that this "oddity" of trying to create a UEFI VM from XO-lite is still present. I'd bet money that this gets fixed before or when the full release is provided. https://xcp-ng.org/forum/topic/11227/boot-new-vm-to-iso-in-xo-lite Workaround is to either click the button provided to download XOA, or install Linux and XO from sources in BIOS mode. Then you have the full XO to work with to bring everything else up the way you want it. I see it as a minor annoyance right now, not really a bug or problem.

@DustinB Hmm - just got done running mem86+ - 4 passes -- all 14 tests. No RAM errors. I wonder the what would cause this error? I'll probably just save config and reinstall. So strange.