Windows 11 Pro 24H2 VM Random Freezes on XCP-NG 8.3
-
Hi everyone,
I'm experiencing a strange issue with a freshly installed Windows 11 Pro 24H2 VM on my XCP-NG 8.3 host. The VM runs fine for some time but then randomly freezes. Here’s what happens:
- The RAM usage graph in XOA freezes at the last recorded value.
- CPU drops to 0% usage.
- RDP disconnects and never comes back.
- XOA console is completely frozen, making interaction impossible.
- The VM still responds to ping (Uptime Kuma reports no ICMP downtime).
- Trying to connect via RDP results in a loading screen for several minutes, but eventually fails.
- It looks like the VM goes into some weird sleep/hibernation mode and won't wake up.
- The only solution is a forced reboot via XOA/XCP-ng Center.
- The Xen Tools are installed and up to date.
Does anyone have any idea what might be causing this? I'm completely stuck...
Here’s a graph showing the issue:
Thanks in advance for your help!
-
@DevilDan What if you uninstalled all Xen tools?
-
@dinhngtu Thank you! I will try uninstalling the Xen Tools and keep you updated.
Here’s some additional information:
- The VM remains in "running" mode, still showing as green.
- I left the VM in its "frozen" state to monitor if it would eventually lose ping or RDP connectivity via Uptime Kuma.
- Here’s the timestamped log of the disconnections:
- RDP (TCP 3389) disconnected around 00:00 - 01:00 (French time).
- Uptime Kuma detected TCP 3389 as down at 02:09.
- Ping was lost at 08:44 (French time) according to Uptime Kuma.
Here’s the corresponding graph:
I'll keep you posted on the results after uninstalling the Xen Tools!
-
@DevilDan Could you also describe your host hardware and special configuration (if any)? Does it happen on other operating systems or just Windows 11?
-
Server: So You Start SYS-1
- CPU: Intel Xeon-E 2136 (6 cores / 12 threads, 3.3 GHz / 4.5 GHz boost)
- RAM: 64GB DDR4 ECC
- Storage: 2x 4TB HDD (Soft RAID)
- Bandwidth: 500 Mbps to 1 Gbps (guaranteed)
- Network: Physical Intel NIC (exact model unknown)
VM Configuration:
- Windows 11 Pro 24H2
- vTPM and Secure Boot enabled (installation worked fine)
- Installed with a local account (no Microsoft account)
- Network:
- The VM is on a subnet managed by OPNSense, which runs as another VM.
- The XCP-NG host has a WAN IP, and OPNSense has both a WAN IP and an additional Failover IP (FO IP).
- The Windows VM uses an Intel E1000 virtual NIC in XOA instead of the default Realtek option.
System Configuration Summary (XCP-NG Server)
Network Interface
- Model: Intel Ethernet Controller X550
- Number of interfaces: 2 (eth0 and eth1)
- Supported speeds: 100M, 1G, 2.5G, 5G, 10G
- Current speed: 1G Full Duplex
Processor
- Architecture: x86_64
- Total cores: 12 (6 physical cores, 2 threads per core)
- Clock speed: ~3.4 GHz
Memory
- Total memory: ~64 GB
- Free memory: ~20 GB
Storage
- Detected disks: 2x HGST HUS726T4TAL (4 TB each)
- Other disks present but not detailed
Network Configuration (XCP-NG)
- Interface eth0 → Assigned to WAN
- Interface eth1 → Internal Pool
Before uninstalling :
-
@dinhngtu said in Windows 11 Pro 24H2 VM Random Freezes on XCP-NG 8.3:
@DevilDan Could you also describe your host hardware and special configuration (if any)? Does it happen on other operating systems or just Windows 11?
Sorry, I forget to answer to the question, it's happen only with W11 pro 24H2, I have another XCP-NG 8.2, with W10 VM it is ok with that WM on the other server.
-
Hi @dinhngtu , it is frozen
EDIT :
I'm installing prometheus and grafana and winsows exporter to have date, time and metrics for CPU, when freeze arrive, exact date and time, to coroborate logsThere is a new update on 24H2, correct RDP bugs, maybe freeze bug ? I test and i'm back if i have more informations :
https://www.it-connect.fr/windows-11-kb5052093-quels-sont-les-changements-apportes-par-cette-mise-a-jour/"Il y a deux bugs corrigés dans la fonctionnalité de Bureau à distance, notamment où l'application cesse de répondre, et un autre bug lié à des problèmes d'affichage."