Proxmox node keeps rebooting

[aryan.sx]

Firstly I would like to preface that I am a complete beginner to proxmox. I have 2 Hp mini pcs running proxmox clustered together. For some reason the first node keeps reboot after about 2 minutes and I'm not sure why. The end of the system logs always seem to output this right before rebooting. The end of the logs vary sometimes but I'm unable to deduce the problem.

Apr 07 21:50:54 pve1 chronyd[813]: System clock TAI offset set to 37 seconds
Apr 07 21:50:54 pve1 pvescheduler[1139]: starting server
Apr 07 21:50:54 pve1 systemd[1]: Started pvescheduler.service - Proxmox VE scheduler.
Apr 07 21:50:54 pve1 systemd[1]: Reached target multi-user.target - Multi-User System.
Apr 07 21:50:54 pve1 systemd[1]: Reached target graphical.target - Graphical Interface.
Apr 07 21:50:54 pve1 systemd[1]: Startup finished in 4.928s (firmware) + 11.043s (loader) + 2.786s (kernel) + 13.780s (userspace) = 32.540s.
Apr 07 21:50:55 pve1 chronyd[813]: Selected source 99.28.14.242 (2.debian.pool.ntp.org)
Apr 07 21:50:59 pve1 pvedaemon[1100]: <root@pam> successful auth for user 'root@pam'
Apr 07 21:51:13 pve1 netavark[715]: timeout met: exiting after 30 secs of inactivity
Apr 07 21:51:13 pve1 systemd[1]: netavark-dhcp-proxy.service: Deactivated successfully.

 

[SteveITS]

With or without a Qdevice for a third vote?

What does “pvecm status” show, on both? And “corosync-cfgtool -s”

 

[aryan.sx]

If by Qdevice you mean another node for qourum then I don't have one but plan on setting one up soon with potentially another pc that I have or a spare Pi 2 W.

Node 1:

Cluster information
-------------------
Name:             pvecluster
Config Version:   2
Transport:        knet
Secure auth:      on

Quorum information
------------------
Date:             Tue Apr  7 22:53:07 2026
Quorum provider:  corosync_votequorum
Nodes:            2
Node ID:          0x00000001
Ring ID:          1.364
Quorate:          Yes

Votequorum information
----------------------
Expected votes:   2
Highest expected: 2
Total votes:      2
Quorum:           2 
Flags:            Quorate

Membership information
----------------------
    Nodeid      Votes Name
0x00000001          1 192.168.1.10 (local)
0x00000002          1 192.168.1.11

Local node ID 1, transport knet
LINK ID 0 udp
        addr    = 192.168.1.10
        status:
                nodeid:          1:     localhost
                nodeid:          2:     connected

Node 2:

Cluster information
-------------------
Name:             pvecluster
Config Version:   2
Transport:        knet
Secure auth:      on

Quorum information
------------------
Date:             Tue Apr  7 22:55:32 2026
Quorum provider:  corosync_votequorum
Nodes:            2
Node ID:          0x00000002
Ring ID:          1.36d
Quorate:          Yes

Votequorum information
----------------------
Expected votes:   2
Highest expected: 2
Total votes:      2
Quorum:           2 
Flags:            Quorate

Membership information
----------------------
    Nodeid      Votes Name
0x00000001          1 192.168.1.10
0x00000002          1 192.168.1.11 (local)

Local node ID 2, transport knet
LINK ID 0 udp
        addr    = 192.168.1.11
        status:
                nodeid:          1:     disconnected
                nodeid:          2:     localhost

I was able to get these while the first node was online before it reboots again

 

[cheiss]

Apr 07 21:51:13 pve1 netavark[715]: timeout met: exiting after 30 secs of inactivity
Apr 07 21:51:13 pve1 systemd[1]: netavark-dhcp-proxy.service: Deactivated successfully.

These are not services shipped with Proxmox VE. Did you install any additional packages and/or services on the host? Have you checked whether they might be interfering?

Also, with 2 nodes you essentially need an external vote aka. QDevice and should never do clustering without. See also Corosync External Vote Support on why.

 

[SteveITS]

status:
nodeid: 1: disconnected

A cluster node will fence and reboot if it can't communicate with the others. With 2 available votes, 1/2 is 50% which is not over 50%, so there is no quorum if one node is offline. You might check firewall rules are allowing corosync from the other node. Otherwise it looks like there's some communication problem between them.

 

[aryan.sx]

These are not services shipped with Proxmox VE. Did you install any additional packages and/or services on the host? Have you checked whether they might be interfering?

The only ran the "PVE Post Install" script from the community-scripts.org website, an ubuntu lts server vm, and pihole. I will try and disable this service while it is online to see if that fixes the issue.

Also, with 2 nodes you essentially need an external vote aka. QDevice and should never do clustering without. See also Corosync External Vote Support on why.

A cluster node will fence and reboot if it can't communicate with the others. With 2 available votes, 1/2 is 50% which is not over 50%, so there is no quorum if one node is offline. You might check firewall rules are allowing corosync from the other node. Otherwise it looks like there's some communication problem between them.

I will setup another node today and see if that also fixes the problem.

Thank you both for the help, I really appreciate it.