Healthcheck for a proxmox server configuration?

[DanH]

I fear that I messed up my server with a failed clustering attempt.
I removed all cluster remainingings now. All is working fine again on the server. I cannot see any issue in any logs.

One VM (Windows 2022) seems to have issues, though. But I cannot pinpoint them. Seems related to networking, but all networking tests are positive.

I wonder if there is some kind of healthcheck I could run. Something that checks the configuration files (i.e. in /etc/pve/*), perhaps file permissions, network settings and others.

I do have a backup of /etc/pve/* and some other files, including /etc/network/interfaces.
I will diff them to the active files.

Is there something else I can do?
Perhaps some genius has created a "check-proxmox-health" script, or such?

Dan

 

[UdoB]

Perhaps some genius has created a "check-proxmox-health" script, or such?

Actually that very term gives several hits: https://www.startpage.com/sp/search?query=check-proxmox-health

Unfortunately I am not aware of a generic, one-script-fits-all solution for your idea. The reason is obvious: there are more combinations of the hardware stack plus the used storage approach plus network stack plus xyz that you would need to check for more combinations than there are stars in the sky.

Some monitoring systems include a Proxmox Template, for example https://www.zabbix.com/integrations/proxmox but that's probably not what you are looking for.

Interestingly the "upgrade checker script" "pve7to8" can run while already on version 8 and it checks for some basic problems.

 

[DanH]

pve7to8 is a good hint.
No unexpected error here, just a warning on the running VM and some updates not yet installed.
(I am somewhat afraid to update libpve-rs-perl, libpve-storage-perl, qemu-server, pve-manager because I messed up the cluster join/remove)

To explain my problem. Perhaps you have another idea.

That one Win2022 VM is hosting a video surveillance system. It gets camera feeds from an AXIS P7316 box that basically translates analog coax camera signals into TCP/IP streams. Those are sent to a service on the VM and then fed into the surveillance system.

I had to hardreboot the proxmox server where this VM is running on, because I messed up adding a cluster member. After following some (mis?)guides to set the system back to a single node, I lost the content of /etc/pve. I had a backup but was not able to restore that because of the corrupt(?) cluster filesystem /etc/pve (even after unmount, it was emtpy). After I lost all access to any proxmox function, I rebooted the system.

After reboot, the proxmox server was back in single node status and all content of /etc/pve was here again.
All configuration of the server seem ok.

The VM is running as well without a problem discovered, until I found out that the surveillance system did not have any active camera feeds anymore.
I was checking all I could, also with a technician from the company which installed the system. No luck.

I restored the VM from a backup where the surveillance system was running smoothly. Still no camera feeds.
We restored the surveillance system configuration. Still no camery feed.

The network is ok as far as I can tell. According to the technician, the system should work.

I controlled the proxmox network settings again and found no error.

For us, it looks as if the camera feeds are just not connected.

Any idea what could be wrong?

 

[UdoB]

Any idea what could be wrong?

Not specifically for that case, sorry.

The VM is fetching data from AXIS. I am not in Windows network debugging, but there should be a tool to check if AXIS is reachable from Windows. It the surveillance software from a third party software vendor? If yes: look for a tool from AXIS to check connectivity. Can Windows just "ping" that AXIS source?

Good luck!