[SOLVED] Web GUI Inaccessible to All but 1 Computer

A

aloanwolf

New Member
Feb 3, 2026
8
0
1
I have a standalone PVE server that has been working fine for the last week or so. I just tried to log into the web GUI from another device, and it kept telling me my password was wrong (I was doing a copy/paste) and eventually the web GUI and SSH just started refusing connection attempts from the secondary device, and any subsequent new devices, but it's still working on the first machine I'd used to set it up in the first place. I can log back in on the first device, using the same clipboard copy/paste that had denied me before. I did confirm it was using PAM, as well.

I've restarted the pveproxy and pvedaemon services and the behavior persists. I don't have Fail2Ban, and the firewalls aren't enabled, so I'm not sure where my connections are getting blocked from. PVE is on a static IP outside of the DHCP scope, so I don't believe it's any sort of IP conflict. I haven't tried rebooting the server yet, but I'd like to know what happened and why, so I can avoid potentially locking myself out in the future.
 
Have you tried to login from physical console with the same password? Are you sure you don't have two PVE nodes running by accident?
Since you already tried to restart pveproxy, try to stop it - did the GUI stop responding for all hosts?
When you are trying to SSH - monitor the log from the "good" host, do you see any connection attempts?

Vanilla Proxmox does not have anything built-in that would block your SSH or GUI connections. Retrace your steps , check history - standard PVE does not behave this way.

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
 
bbgeek17 said:
Have you tried to login from physical console with the same password? Are you sure you don't have two PVE nodes running by accident?
Since you already tried to restart pveproxy, try to stop it - did the GUI stop responding for all hosts?
When you are trying to SSH - monitor the log from the "good" host, do you see any connection attempts?

Vanilla Proxmox does not have anything built-in that would block your SSH or GUI connections. Retrace your steps , check history - standard PVE does not behave this way.

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
Click to expand...
The password does work at the console. I'm not sure how I would have another PVE node running accidentally.

Stopping the pveproxy service does cause the GUI to stop responding for the "good" host, and it responds on the same host after restarting the service. I still cannot seem to access it from any other machine; the browser and Putty are giving me "connection refused" errors. Pings are fine, though. I was able to get the web GUI to at least prompt me to log in on the non-good host prior to the login failure. The only steps between accessing the GUI on the non-good host and the refused connection attempts was the failed login attempts that were made.

Which log do I need to check to view the connection attempts? (Also, thank you for your quick response!)
 
aloanwolf said:
The password does work at the console. I'm not sure how I would have another PVE node running accidentally.

Stopping the pveproxy service does cause the GUI to stop responding for the "good" host, and it responds on the same host after restarting the service. I still cannot seem to access it from any other machine; the browser and Putty are giving me "connection refused" errors. Pings are fine, though. I was able to get the web GUI to at least prompt me to log in on the non-good host prior to the login failure. The only steps between accessing the GUI on the non-good host and the refused connection attempts was the failed login attempts that were made.

Which log do I need to check to view the connection attempts? (Also, thank you for your quick response!)
Click to expand...
Quick update: While refreshing on the bad-host, it did give me an unsigned SSL warning, which I accepted, then it gave me a connection refused error again. This was the only time I've seen the SSL warning since getting "locked out", and I'm not able to replicate this behavior again.
 
It does sound very much like a duplicate IP issue.

The logs are : journalctl -f
You can also run tcpdump from the PVE console or SSH via "good" host to see what, if anything, happens to your "bad" connections:
tcpdump -i physicalInterfaceHere -n host bad.client.ip.here

Also, check ARP table and compare MAC addresses for PVE host from bad and good.

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
 
bbgeek17 said:
It does sound very much like a duplicate IP issue.

The logs are : journalctl -f
You can also run tcpdump from the PVE console or SSH via "good" host to see what, if anything, happens to your "bad" connections:
tcpdump -i physicalInterfaceHere -n host bad.client.ip.here

Also, check ARP table and compare MAC addresses for PVE host from bad and good.

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
Click to expand...
Thank you for the suggestion on checking the ARP tables, I did see they were different between the two machines. Flushing the ARP cache didn't seem to help, so I just moved PVE to a new IP and now I can connect on the other machines. The MAC address in the ARP table was another device with a different static IP. Not sure what was going on, as I cannot ping anything at the previous IP address, but satisfied that I shouldn't likely be "locked out" again in the future.

Thanks again for your help, I very much appreciate it!
 
You must log in or register to reply here.
Top Bottom
Back