VNC consoles extremely slow

number5

New Member
Mar 22, 2021
19
1
3
United States
The issue I am seeing is that VNC consoles appear to be very slow, and with ~30 users connected, practically unusable, with keystroke latency reaching upwards of 10 seconds, intermittently but frequently. The VMs, however, are fine. RDPing or SSHing in shows that the VMs are just as responsive as you would expect them to be. I have also tried SPICE on a few of these VMs, and this, for now, seems to work to bypass the issue.

In addition, it does not seem as though restarting pveproxy.service has any impact console performance: the VNC consoles continue to perform poorly—even after the majority of the consoles are closed. Any new console feels sluggish, and even though they become much more responsive, intermittent latency is still present and frequent enough to cause issues with typing. Switching browsers or rebooting the VMs does not appear to resolve the issue either.

The performance issue does not appear to affect the PVE web portal. I am also not seeing any abnormal IO, memory, network, or CPU usage on any of the proxy hosts or any of the PVE hosts.

Has anyone seen a similar issue? Any thoughts about what could be causing this?
 
The issue I am seeing is that VNC consoles appear to be very slow, and with ~30 users connected, practically unusable, with keystroke latency reaching upwards of 10 seconds, intermittently but frequently. The VMs, however, are fine. RDPing or SSHing in shows that the VMs are just as responsive as you would expect them to be. I have also tried SPICE on a few of these VMs, and this, for now, seems to work to bypass the issue.

In addition, it does not seem as though restarting pveproxy.service has any impact console performance: the VNC consoles continue to perform poorly—even after the majority of the consoles are closed. Any new console feels sluggish, and even though they become much more responsive, intermittent latency is still present and frequent enough to cause issues with typing. Switching browsers or rebooting the VMs does not appear to resolve the issue either.

The performance issue does not appear to affect the PVE web portal. I am also not seeing any abnormal IO, memory, network, or CPU usage on any of the proxy hosts or any of the PVE hosts.

Has anyone seen a similar issue? Any thoughts about what could be causing this?
I have some updates regarding the issue. It looks like the web portal was indeed being affected, but only while the users were logged in contemporaneously. Once the users logged off, the web portal performance returned to normal, unlike the VNC consoles. In any case, fixing the web portal performance was quite easy. To do that, I increased the pveproxy (/usr/share/perl5/PVE/Service/pveproxy.pm) max_workers to 20. In my testing, this seemed to be an adequate number to support ~30 users on one web portal.

I am still trying to figure out what the bottleneck is with the VNC consoles.
 
  • Like
Reactions: bobmc

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get yours easily in our online shop.

Buy now!