Proxmox random reboots

L

Laimonas Vijūnas

Active Member
Apr 9, 2019
8
0
41
36
It all started when I updated to 9.2.11 using pve-no-subscription repositories. Proxmox reboots randomly after 10-20 minutes. Can't find anything meaningful in logs, except :
Aug 28 13:12:46 test pvestatd[1820]: storage 'Buffalo' is not online
Aug 28 13:12:47 test pvestatd[1820]: status update time (10.049 seconds)
Aug 28 13:15:49 -- Reboot --
Aug 28 13:15:49 test kernel: Linux version 7.0.14-14-pve (build@proxmox) (gcc (Debian 14.2.0-19) 14.2.0, GNU ld (GNU Binutils for Debian) 2.44) #1 SMP PREEMPT_DYNAMIC PMX 7.0.14-14 (2026-08-22T15:01Z) ()
 
-Did you rule out fencing in case it is a cluster node?
- Does Previous system reset reason shows something interesting?
journalctl -b | grep "system reset reason"

Do you have in IPMI/iDRAC? does it show anything in health event logs/sensor reading?
 
It's not cluster.
journalctl -b | grep "system reset reason" gives no output
It's ASUS motherboard ( if I remember correctly), so sadly no Ipmi/Idrac.
On last reboot found never seen log entry:
Aug 28 13:17:44 test QEMU[2319]: ehci: PERIODIC list base register set while periodic schedule
Aug 28 13:17:44 test QEMU[2319]: is enabled and HC is enabled
But it doesn't look relevant.

Setup specs:
I7-14700K
4x32GB DDR5 RAM
 
What kernel version did work well? Can you still boot that one without issues? If there is a large jump in kernel version then you might need a BIOS update. Otherwise, it might be a hardware problem (like dying PSU) that just happened around the same time.
 
At the moment I am running on 7.0.14-14-pve. I didn't document previous version but I install updates every 4 weeks, so it can't be large jump. Installed updates on Sunday, server started to acting strange on Tuesday. Today I installed latest updates when it started reboot after 10-20 minutes. Didn't help. 1787916303265.png
Working remote so don't have opportunity to boot older kernels for testing. Bios was updated to latest version last year. I'll test this theory when I'll migrate my VMs to another host. Would be crazy luck with dying PSU day after proxmox update, but it's also possible.

A bit scary because sometimes it works fine for hour or more, sometimes it reboots every 3-10 minutes.
 
Impact said:
These are kinda infamous for being unstable. Search for intel degradation.
Click to expand...
Wasn't the "official Intel recommendation" to increase the Dynamic Vcore voltage by 0.005V, reboot and see what happens? :D
 
One thing that may happen during the first 30 minutes is that caches are warmed up, ARC may fill, other processes start and consume memory.

Let memtest86+ run for multiple rounds, at least over night - or longer.
 
You must log in or register to reply here.
Top Bottom
Back