[SOLVED] Problems with PVE webapp no longer being accesible

Balthazar_03

New Member
Jun 4, 2025
3
0
1
Something I've done has made the web UI inaccesible.
I can still ping, and access the server's via SSH, but cannot access the webUI of it.

What I remember doing was trying to update the server, for which I made the mistake of doing apt upgrade, not apt dist-upgrade.

Beyond this I've got a cluster with two nodes (makes it easier to work on both), and I'm trying to switch the IP I'm working with to a different LAN (going from 192.168.0.0 to a private 10.0.0.0)


What I remember is I was trying to prep as much as I could to make the switch as easy as possible, but still left everything functional. Found an update button somewhere in the web-ui, and pressed it. After that my secondary node became unresponsive to the cluster and I could no longer access its web UI.
I then thought it might be an issue with the software being out-of-date and tried updating the primary node. Now I can no longer access either via web-ui.

I may be a little slow to react, due to only having access to the server post 18:00 cest, but I'll try and give any additional information that is needed.
If anyone can help, it'd be much appreciated 'cause I'm at my wits end with this.
 
Hi @Balthazar_03 , welcome to the forum.

When you get a chance, please review threads similar to your situation. They contain asks from the community that help create a full picture (actual data output from the system). It would be great if you can aggregate that and provide along with your question.

etc

Also, please note that having 2-node cluster is not supported and is a recipe for access loss during maintenance. If you plug "two node cluster" into the forum search you will find a lot of information about this.

Good luck


Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
 
Hi @Balthazar_03 , welcome to the forum.

When you get a chance, please review threads similar to your situation. They contain asks from the community that help create a full picture (actual data output from the system). It would be great if you can aggregate that and provide along with your question.

etc

Also, please note that having 2-node cluster is not supported and is a recipe for access loss during maintenance. If you plug "two node cluster" into the forum search you will find a lot of information about this.

Good luck


Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

Thank you for linking these forums!
The actual links and quick answer is very much appreciated <3

I'll look through them when I get the chance.

Again, thanks for the quick answer and guidance!
 
To recap and close this question.
Today (10/06/2025) I had the time and energy to try and fix this.
Like was said by bbgeek17, having a 2-node cluster is a BAD idea.

In my case it caused issues with the pmxcfs process.
From what I understand of ChatGPT's filtering (just ignore the AI. I'm not happy with it, but there's too much information and my brain can't handle or filter it all, and this is the best solution for that.) the pve-cluster service tried to get a lock on the proxmox config, but there was already one (a broken one) meaning it couldn't get it, meaning /etc/pve remained empty and inaccessible.
This then had the effect of the server remaining accesible, but the webui being broken.

To fix it I read what pmxcfs processes were running,
ps aux | grep pmxcfs
killed them,
pkill -9 pmxcfs
And then restarted the cluster
systemctl start pve-cluster

With my other (I think so far unmentioned) server that was running into the same issues, I also removed the cluster in it's entirety.
Something I did somewhere already removed the first from it, so I just had to remove the remnants of the cluster from the secondary server.
systemctl stop pve-cluster corosync
rm -r /etc/pve/corosync.conf
rm -r /etc/corosync/*
rm -r /var/lib/pve-cluster/*
systemctl restart pve-cluster
systemctl status pve-cluster
After this I just had to restore my Vm's and everything was working again.