@XCP-ng-JustGreat I only use the passthrough part of disks, not controllers. VUSB part in xo i don't use. I think that's a completely different approach. Maybe you could check your drivers and manager. There is a cleaner somewhere, don't know the link, to remove the old drivers in windows with powershell and after that install the xen drivers XenServer VM Tools for Windows 9.4.1. I got it from here: https://www.xenserver.com/downloads https://github.com/xcp-ng/win-pv-drivers/releases/download/v9.0.9065/XenClean.zip docs: https://docs.xcp-ng.org/vms/#fully-removing-xen-pv-drivers-with-xenclean

Update: I was able to move the VM to a different host that is also xcp-ng 8.2. When I tried moving the VM back to where it came from now I'm getting: NO_HOSTS_AVAILABLE() vm.migrate { "mapVdisSrs": { "f4d88025-9e26-44a1-8028-4d5d9c11c994 ": "18850841-5eb8-6b28-73ba-cae1c40af157" }, "mapVifsNetworks": { "b56126a8-85ab-80de-9db8-4864ec5618e5 ": "99b40169-8dab-6e9d-db96-23277b405b4a " }, "migrationNetwork": "99b40169-8dab-6e9d-db96-23277b405b4a ", "sr": "18850841-5eb8-6b28-73ba-cae1c40af157", "targetHost": "69c0c257-a547-4d49-8270-b49eafe35c2c", "vm": "78938c32-eb52-491f-1fbf-3ad5d2e374f0" } { "code": "NO_HOSTS_AVAILABLE", "params": [], "task": { "uuid": "4eb56a28-ffa7-cee6-31e1-87209af8de75 ", "name_label": "Async.VM.migrate_send", "name_description": "", "allowed_operations": [], "current_operations": {}, "created": "20250131T20:51:54Z", "finished": "20250131T20:51:54Z", "status": "failure", "resident_on": "OpaqueRef:72cbec6b-d489-49ca-a8a8-e623e3c5f4e4", "progress": 1, "type": "<none/>", "result": "", "error_info": [ "NO_HOSTS_AVAILABLE" ], "other_config": {}, "subtask_of": "OpaqueRef:NULL", "subtasks": [], "backtrace": "(((process xapi)(filename ocaml/xapi/xapi_vm_placement.ml)(line 104))((process xapi)(filename ocaml/xapi/message_forwarding.ml)(line 1199))((process xapi)(filename lib/xapi-stdext-pervasives/pervasiveext.ml)(line 24))((process xapi)(filename lib/xapi-stdext-pervasives/pervasiveext.ml)(line 35))((process xapi)(filename ocaml/xapi/helpers.ml)(line 1356))((process xapi)(filename ocaml/xapi/message_forwarding.ml)(line 1191))((process xapi)(filename ocaml/xapi/message_forwarding.ml)(line 2279))((process xapi)(filename lib/xapi-stdext-pervasives/pervasiveext.ml)(line 24))((process xapi)(filename ocaml/xapi/rbac.ml)(line 205))((process xapi)(filename ocaml/xapi/server_helpers.ml)(line 95)))" }, "message": "NO_HOSTS_AVAILABLE()", "name": "XapiError", "stack": "XapiError: NO_HOSTS_AVAILABLE() at Function.wrap (file:///opt/xo/xo-builds/xen-orchestra-202501311655/packages/xen-api/_XapiError.mjs:16:12) at default (file:///opt/xo/xo-builds/xen-orchestra-202501311655/packages/xen-api/_getTaskResult.mjs:13:29) at Xapi._addRecordToCache (file:///opt/xo/xo-builds/xen-orchestra-202501311655/packages/xen-api/index.mjs:1068:24) at file:///opt/xo/xo-builds/xen-orchestra-202501311655/packages/xen-api/index.mjs:1102:14 at Array.forEach (<anonymous>) at Xapi._processEvents (file:///opt/xo/xo-builds/xen-orchestra-202501311655/packages/xen-api/index.mjs:1092:12) at Xapi._watchEvents (file:///opt/xo/xo-builds/xen-orchestra-202501311655/packages/xen-api/index.mjs:1265:14)" } I rebooted XO and tried moving the VM to the xcp-ng 8.3 machine again. It looks like the task started but now it's hung at "Connection to VM console (on xcp-m83) 0%" and I can't "Cancel" or "Destroy" the task

No specific plans as this kind of feature isn't really useful in Datacenters. However, if there's a way that's relatively universal, it's doable with the help of the community (ie contributions are very welcome and we'll review it seriously)

@apz I see. Thanks.

@andrewperry The release notes warn about PV guests.

In any case, you can ignore it.

@Kajetan321 You should consider the effects of NUMA/vNUMA in terms of performance. Crossing over to other physical CPUs or memory banks will create slowdowns. If possible, unless you absolutely need all those VCPUs on any or both VMs, you may be best off splitting the 36 VCPUs and assigning 18 on each VM, checking to see if all the memory stays with the bank(s) of those specific VCPUs. See my articles on the CUCG site about NUMA and in particular, graphics performance (and of course, calculations will also be affected). Best would be to run benchmarks with both configurations. Shrinking the number of VCPUs on a running VM can lead to issues (as indicated in the video posted just before my response). See the Tale of Two Servers articles here for details: https://community.citrix.com/profile/46909-tobias-kreidl/#_=_

Pinging @florent about this

Ghost XO user authentication task has nothing to do with backup restore, it is a known bug that will likely be fixed in the next release.

Hrm, I've always just reattached to the backup repo so I could see everything. Must've been a misunderstanding on my part. That's what I meant. Even a fresh XO without any previous config could restore backup if you add the backup repository at the same path where you had your backups before Sure, it's even easier to restore XO metadata because you don't have to remember where are you backups, but still, worst case scenario (losing everything except the BR), it's possible to restore.

I am using USB-servers from SEH and Silex. Both of them are working. Silex can provide higher Performance - SEH is IMHO nicer for license-dongels, etc.

It's important to understand the components you are using, especially on a software that you use without support/training In short: SR: a Storage Repository is access by your host/dom0/XCP-ng, and will be used to store ISOs and/or VM disks. It's running on a relatively old CentOS release (7.x), and will mount with some specific settings BR/remote: a Backup Repository where Xen Orchestra will send the backup of your VMs. Usually, Xen Orchestra runs (in XOA) inside a Debian 11 VM, with default NFS mount options (default: ie from the operating system point of view). It's hard to tell exactly the issue since you are likely using XO from the sources, so a XO that you choose to install on a system where we (aka "XO devs") have precisely 0 control on. Plus the NFS configuration part is also out our scope because it's not running inside our infrastructure too. That's a lot of moving pieces, that's why it's important to understand what you do so we can assist here the most efficient way we can

@Kajetan321 No static memory would be 28G/28G 8/28 is dynamic memory

@olivierlambert Thank you for following up.

@olivierlambert Yes, each job has a unique identifier and willl not touch backup files of other jobs.

@stormi Yeah, seems better this way ^^ I will try to write a draft in the next few days to update the documentation

Sadly, it's hard to know what happened without getting the errors you had (eg when you tell: "VM will not start" is very vague).

@Darkbeldin Got ya, thanks!

@olivierlambert Thanks!