I'm getting frequent gaps in my summary graphs and trying to pin point the problem and see if it is perhaps affecting other features.
Anyone seen (and maybe figured out) the weird charting in the picture? I run a single node setup on a Dell T620; networking wise the host takes DHCP from a glinet opal router which is a wireguard client to a edge router lite server using split tunneling.
Update:
Ok, I think this is a symptom to a network problem and has affected a number of users, both prosumer and big multi-node setups.
Symptoms:
- gaps in the performance summary
- pvesm takes forever to load (minutes instead of seconds)
- client vanishing in logs
- slow loading in web gui, especially with storage, until a communication error comes up
Things it could be:
- Custom/old certificate gone awry: can recreate and restart all services. This did not fix it for me.
https://forum.proxmox.com/threads/web-gui-communication-failure-for-all-components.41233/
- udisks2: this tends to cause lots of high CPU usage problems, I was desperate and tried masking it. pvesm still slow.
- jumbo frame sizes: ive only used defaults
- cifs mount via cli: i did not use the web GUI to set this up and also have a other mounting issues. But this timing out could be affecting everything else.
I'll do a series of reboots (which magically fixes issues for many) and then post some logs if it persists.
Anyone seen (and maybe figured out) the weird charting in the picture? I run a single node setup on a Dell T620; networking wise the host takes DHCP from a glinet opal router which is a wireguard client to a edge router lite server using split tunneling.
Update:
Ok, I think this is a symptom to a network problem and has affected a number of users, both prosumer and big multi-node setups.
Symptoms:
- gaps in the performance summary
- pvesm takes forever to load (minutes instead of seconds)
- client vanishing in logs
- slow loading in web gui, especially with storage, until a communication error comes up
Things it could be:
- Custom/old certificate gone awry: can recreate and restart all services. This did not fix it for me.
https://forum.proxmox.com/threads/web-gui-communication-failure-for-all-components.41233/
- udisks2: this tends to cause lots of high CPU usage problems, I was desperate and tried masking it. pvesm still slow.
- jumbo frame sizes: ive only used defaults
- cifs mount via cli: i did not use the web GUI to set this up and also have a other mounting issues. But this timing out could be affecting everything else.
I'll do a series of reboots (which magically fixes issues for many) and then post some logs if it persists.
Attachments
Last edited: