[SOLVED] NoVNC not working after upgrade from 8 to 9

bibear

New Member
May 6, 2026
2
0
1
Hi!

we upgraded one of our pve servers from 8 to 9, and now cannot access the VMs anymore via VNC console. Gui just says "timeout", in the log theres following:

Code:
May 06 10:33:08 vmhost2 pvedaemon[692919]: starting vnc proxy UPID:vmhost2:000A92B7:00B8D5EF:69FAFCC4:vncproxy:100:user@pam:
May 06 10:33:08 vmhost2 pvedaemon[689701]: <user@pam> starting task UPID:vmhost2:000A92B7:00B8D5EF:69FAFCC4:vncproxy:100:user@pam:
May 06 10:33:08 vmhost2 pveproxy[687303]: connect to 'localhost:5900' failed: Connection refused at /usr/share/perl5/PVE/APIServer/AnyEvent.pm line 579.
May 06 10:33:13 vmhost2 pveproxy[687303]: worker exit
May 06 10:33:13 vmhost2 pveproxy[17064]: worker 687303 finished
May 06 10:33:13 vmhost2 pveproxy[17064]: starting 1 worker(s)
May 06 10:33:13 vmhost2 pveproxy[17064]: worker 692930 started
May 06 10:33:18 vmhost2 pvedaemon[692919]: connection timed out
May 06 10:33:18 vmhost2 pvedaemon[689701]: <user@pam> end task UPID:vmhost2:000A92B7:00B8D5EF:69FAFCC4:vncproxy:100:user@pam: connection timed out

Thanks in advance for any help!
 
hi!

could you please run `ss -ltapn | grep 5900` exactly in those 10 seconds between "starting vnc proxy" and the "connection timed out" and post the result here, together with `ip a` (if you censor the output, please do so in a way that does not lose important data like ipv4 vs ipv6)? thanks!
 
  • Like
Reactions: fiona
Hi!

thanks, i just found the problem. In my /etc/hosts existed a line with the external ipv6 address and name localhost, tcpdump on lo0 showed me that it tried to connect to localhost port 5900, but ss -latpn showed vnc listening on the external v6 address.
Removed the line in /etc/hosts, restarted pveproxy, now it works :)