Okay, so... this is kernel version specific = it's a kernel issue. Under 6.14.11-3, the problem is there. Booting with the 6.14.8-2 kernel, I've had no 'oops' at all.
I have two proxmox servers, the other is okay with 6.14.11-3 but may not be...
Okay, so it's not sysstat causing it directly, it died again with that disabled. It's also not memory - left memtest running for 4 full passes and it passed 100% cleanly.
Looking at the stack trace in the log I shared, the panic appears to have...
One of my PVE hosts hung today. Web UI alive and stats updating, but no actions possible, no access to VM/LXC consoles or host shell. Ping response, SSH not. No obvious issues in the stats but they were pretty much flat for a couple of hours...