Unexpected reboot of Proxmox host

[Merowing]

Hi Proxmox community,

we are new using Proxmox and need your help now. Unfortunately we had an unexpected Proxmox host reboot.
Here are the facts:
- Proxmox Version 9.1.4
- Host is in a cluster with 2 others, HA worked
- Reboot was Feb 10, 08:18:18 - 08:18:21

System log:
Feb 10 08:18:15 opspsrv513 sshd-session[3963687]: Accepted password for root from 193.141.181.62 port 59590 ssh2
Feb 10 08:18:15 opspsrv513 sshd-session[3963687]: pam_unix(sshd:session): session opened for user root(uid=0) by root(uid=0)
Feb 10 08:18:15 opspsrv513 systemd[1]: Created slice user-0.slice - User Slice of UID 0.
Feb 10 08:18:15 opspsrv513 systemd[1]: Starting user-runtime-dir@0.service - User Runtime Directory /run/user/0...
Feb 10 08:18:15 opspsrv513 systemd-logind[2529]: New session 40233 of user root.
Feb 10 08:18:15 opspsrv513 systemd[1]: Finished user-runtime-dir@0.service - User Runtime Directory /run/user/0.
Feb 10 08:18:15 opspsrv513 systemd[1]: Starting user@0.service - User Manager for UID 0...
Feb 10 08:18:15 opspsrv513 (systemd)[3963693]: pam_unix(systemd-user:session): session opened for user root(uid=0) by root(uid=0)
Feb 10 08:18:15 opspsrv513 systemd-logind[2529]: New session 40234 of user root.
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Queued start job for default target default.target.
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Created slice app.slice - User Application Slice.
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Reached target paths.target - Paths.
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Reached target timers.target - Timers.
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Listening on dirmngr.socket - GnuPG network certificate management daemon.
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Listening on gpg-agent-browser.socket - GnuPG cryptographic agent and passphrase cache (access for web browsers).
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Listening on gpg-agent-extra.socket - GnuPG cryptographic agent and passphrase cache (restricted).
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Starting gpg-agent-ssh.socket - GnuPG cryptographic agent (ssh-agent emulation)...
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Starting gpg-agent.socket - GnuPG cryptographic agent and passphrase cache...
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Listening on keyboxd.socket - GnuPG public key management service.
Feb 10 08:18:17 opspsrv513 systemd[3963693]: Starting ssh-agent.socket - OpenSSH Agent socket...
Feb 10 08:18:18 opspsrv513 systemd[3963693]: Listening on gpg-agent-ssh.socket - GnuPG cryptographic agent (ssh-agent emulation).
Feb 10 08:18:18 opspsrv513 systemd[3963693]: Listening on gpg-agent.socket - GnuPG cryptographic agent and passphrase cache.
Feb 10 08:18:18 opspsrv513 systemd[3963693]: Listening on ssh-agent.socket - OpenSSH Agent socket.
Feb 10 08:18:18 opspsrv513 systemd[3963693]: Reached target sockets.target - Sockets.
Feb 10 08:18:18 opspsrv513 systemd[3963693]: Reached target basic.target - Basic System.
Feb 10 08:18:18 opspsrv513 systemd[3963693]: Reached target default.target - Main User Target.
Feb 10 08:18:18 opspsrv513 systemd[3963693]: Startup finished in 2.334s.
Feb 10 08:18:18 opspsrv513 systemd[1]: Started user@0.service - User Manager for UID 0.
Feb 10 08:18:18 opspsrv513 systemd[1]: Started session-40233.scope - Session 40233 of User root.
Feb 10 08:18:18 opspsrv513 sshd-session[3963719]: Received disconnect from 193.141.181.62 port 59590:11: Connection terminated by the client.
Feb 10 08:18:18 opspsrv513 sshd-session[3963719]: Disconnected from user root 193.141.181.62 port 59590
Feb 10 08:18:18 opspsrv513 sshd-session[3963687]: pam_unix(sshd:session): session closed for user root
Feb 10 08:18:18 opspsrv513 systemd[1]: session-40233.scope: Deactivated successfully.
Feb 10 08:18:18 opspsrv513 systemd-logind[2529]: Session 40233 logged out. Waiting for processes to exit.
Feb 10 08:18:18 opspsrv513 systemd-logind[2529]: Removed session 40233.
-- Reboot --
Feb 10 08:24:21 opspsrv513 kernel: Linux version 6.17.4-2-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 6.17.4-2 (2025-12-19T07:49Z) ()
Feb 10 08:24:21 opspsrv513 kernel: Command line: BOOT_IMAGE=/boot/vmlinuz-6.17.4-2-pve root=/dev/mapper/pve-root ro quiet



Let me know what else you need to helpe me troubleshot this.

Best regards,
Christian

 

[m.j]

Could you send a bit more of the logs so we can get bit more of an idea of what happened? For example, the following could be helpful

journalctl --since="2026-02-10 08:10" --until="2026-02-10 08:30" | gzip > $(hostname)-$(date -Is)-journal.txt.gz

 

[ghusson]

Maybe you are in the case "it's not a bug, it's a feature"
If you use HA, be sure to understand the STONITH mecanism. A visual summary : https://www.virtualizationhowto.com...ter-state-quorum-and-node-majority-1.png.webp
Look at Proxmox Cluster State.
And you can read this : https://pve.proxmox.com/wiki/High_Availability#ha_manager_fencing

 

[UdoB]

STONITH

Just for clarity, while the result may look the same, PVE HA is the opposite: not "shoot-the-other-node and live" but "commit suicide" - a node fences itself, by a hard reset.

 

[Merowing]

journalctl --since="2026-02-10 08:10" --until="2026-02-10 08:30" | gzip > $(hostname)-$(date -Is)-journal.txt.gz

Thanks for your help. Here is the output.

 

[ghusson]

Just for clarity, while the result may look the same, PVE HA is the opposite: not "shoot-the-other-node and live" but "commit suicide" - a node fences itself, by a hard reset.

Oh yes, indeed. thank you for this clarification.

 

[ghusson]

Thanks for your help. Here is the output.

There is no evidence of quorum loss. Corosync does not complains. This there is no a HA protection mecanism involved.
Does it happens regularly ?

 

[Merowing]

There is no evidence of quorum loss. Corosync does not complains. This there is no a HA protection mecanism involved.
Does it happens regularly ?

Well I had 3 servers installed and running for like 3 weeks now. First time it happened. But I need to know why it happend to prevent it in the future cause we are about to run productive VMs on Proxmox soon.

 

[ghusson]

Well I had 3 servers installed and running for like 3 weeks now. First time it happened. But I need to know why it happend to prevent it in the future cause we are about to run productive VMs on Proxmox soon.

I understand but it is very difficult to put a diagnostic on this. I saw no evidence in the logs. Maybe somebody will have an idea ?
What is the hardware of your servers ?

 

[Merowing]

I understand but it is very difficult to put a diagnostic on this. I saw no evidence in the logs. Maybe somebody will have an idea ?
What is the hardware of your servers ?

Hardware is Cisco UCS B200 M5

 

[ghusson]

Ok.
What did you do as special parameters or installs on the servers that doesn't come from PVE ISO installer ?
Did the server already worked on an other OS with any problem ?
Did you look at hardware logs for over-temperature or manual hard reset ?

 

[Merowing]

Ok.
What did you do as special parameters or installs on the servers that doesn't come from PVE ISO installer ?
Did the server already worked on an other OS with any problem ?
Did you look at hardware logs for over-temperature or manual hard reset ?

Nothing special during install. Afterwards chrony to add NTP Server, Adding FC Storage and multipathing.
The server worked with VMware for years without issues.
Hardware Logs do not show anything. No hard reset, no temp, no hardware errors.

 

[ghusson]

Hmmm very stange...
I invite you to configure remote logging (for having fresher info if reboots occurs again).
And maybe to subscribe official support from Proxmox.
It happened only one time ?

 

[m.j]

Since there is not really anything in the logs, and it only happened once, it is really hard to say. Maybe it could be a firmware thing? For example, it seems your BIOS is from 2022.

You could try installing an older kernel (than 6.17.4-2) and then pin it, reboot, and watch the logs if it does it again.