Did a reload of the daemon and log still says:
qemu-system-x86_64: Issue while setting TUNSETSTEERINGEBPF: Invalid argument with fd: 50, prog_fd: -1
qemu-system-x86_64: Issue while setting TUNSETSTEERINGEBPF: Invalid argument with fd: 51, prog_fd: -1
@tomp my kernel version is 5.15.0-73. And i have an epyc 2 cpu. You think i should try the modprobe workaround?
Do those messages appear at first boot or only after crash. They may not be relevant.
After first boot. I booted up. Reloaded the lxd daemon. And the msg was there.
I think we need to find out what process is causing the CPU/load spikes, as I think
TUNSETSTEERINGEBPF is not relevant here. Especially as its occurring at startup when the VMs are working.
The database errors suggest the system is getting overloaded somehow.
Alright. Whats the best way to measure? I have zabbix that has logs on lots of stuff. Also upgraded ram from 128 to 512gb
Update: 26 day uptime now. The win vm has only been onlinefor short periods.