Random host freeze after 60–80h uptime (Intel N150, Proxmox VE 8.x / kernel 6.14) – no logs, full system hang

[romainrossi.osteo]

Hello,
I’m experiencing random full host freezes on a Proxmox VE system.
The freeze occurs between ~60h and 80h of uptime, with no logs, no kernel panic, no reboot, and the system becomes completely unresponsive (SSH, Web UI, VM, console → dead).
Only a hard power cycle recovers the system.
️ Hardware
Intel N150 (Alder Lake-N)
4 cores
16 GB RAM
Single M.2 SATA SSD
No GPU passthrough
No USB passthrough (USB autosuspend disabled)
Fan OK, temps stable (~60–70 °C under load)
Proxmox / Host
Proxmox VE 8.x
Kernel: 6.14.11-4-pve
CPU governor: performance
Ethernet driver: r8169
IRQ pinned to CPU0
Network watchdog enabled
Static IP (no DHCP renew)
No ZFS (ext4)
No ballooning
No host swap activity during freeze
BIOS / GRUB / Power management
Wi-Fi & Bluetooth disabled in BIOS
Intel DTT disabled
ASPM disabled
USB autosuspend disabled
C-States limited to C3
GRUB parameters:
Copier le code

intel_idle.max_cstate=3
processor.max_cstate=3
usbcore.autosuspend=-1
pcie_aspm=off
CPU frequency locked via governor (no scaling)
SSD power management set to performance
️ VM configuration
Home Assistant OS VM
4 vCPU
10 GB RAM
CPU type: host
No PCI / USB passthrough
Stable inside VM (freeze is host-level)
Observations
Memory usage appears stable (no leak visible)
journalctl -k shows nothing before freeze
No OOM killer
No thermal throttling
No watchdog reset
System can run perfectly fine up to ~80h, never more
️ Temporary mitigation
As a workaround, I implemented a conditional uptime-based reboot script (host-side), but this is clearly not a real fix.
Question
Has anyone seen similar freezes on Intel N100 / N150 / Alder Lake-N with recent Proxmox kernels?
Known issues with C-states / idle / PM?
Kernel regression in 6.14.x?
Any additional debug flags or known fixes?
Any feedback or similar experience would be greatly appreciated.
Thanks.

 

[fstrankowski]

Cant you just type your question like every other human beeing and stop wasting this forums people resources with unformatted AI Slop?!

 

[romainrossi.osteo]

Cant you just type your question like every other human beeing and stop wasting this forums people resources with unformatted AI Slop?!

Sorry about the formatting.

Short version: Proxmox VE 8.x host on Intel N150 freezes completely after ~60–80h uptime, no logs, no kernel panic.
It looks related to power management / C-states on Alder Lake-N.

If anyone has seen this or has a lead, I’d appreciate it.

 

[toasun]

I got the same error too.
But the intersting thing is, the virtual machines are still running when the ssh/web/console dead.

 

[hille]

Maybe its connected, but my proxmox host froze fully like a few minutes ago with not even registering keyboard input from a physically connected keyboard (not even sysrq). After pulling the cords from my server I cant find anything in the logs which could be connect to it / no logging at all. But I've still got free space on my dives.

 

[Impact]

Have you installed the microcode package? https://pve.proxmox.com/pve-docs/pve-admin-guide.html#sysadmin_firmware_cpu