Hey there, i have the same problem. The vm wich freezes is a opnsense. I would be so thankful if someone finds a solution.
VM freezes irregularly
- Thread starter thimplicity
- Start date
Can you let us know if your VMs freeze during normal operation (not rebooting) on 5.19.3? Try to keep the VMs running as long as you can without rebooting.
If you have stability over a week, this is a good sign and I might try and repackage the 5.19.3 kernel for testing on Proxmox.
If anyone's interested in running the 5.19.3 mainline kernel on Proxmox, I've repackaged the deb files and removed the zst compression. You can download the repackaged kernel here:
Edit: link removed - there's newer Proxmox kernels available in the post below.
Edit: link removed - there's newer Proxmox kernels available in the post below.
Last edited:
I've been running my host with Proxmox Edge kernel 5.19.4 for almost two days with no crashes, whereas with any other kernel my OPNSense and Home Assistant VMs would crash every 4-5 hours. Here's the link if you want to try for yourselves:
https://github.com/fabianishere/pve-edge-kernel
https://github.com/fabianishere/pve-edge-kernel
I had no idea these kernels existed. It would have been nice to have known when we were troubleshooting.
Are you running the edge kernel? I'll try switching back to Proxmox today and run the edge kernel too.
I did the updates from Proxmox, which offered me the installation of PVE 5.15..39-4.
Since the 2 test VMs are running correctly and quite well I must say...
But I would like it to run for at least 1 week to be able to decide...
ut in any case, for the moment, I touch wood, it is promising...
Since the 2 test VMs are running correctly and quite well I must say...
But I would like it to run for at least 1 week to be able to decide...
ut in any case, for the moment, I touch wood, it is promising...
OK, so I've moved my VMs back to Proxmox and I've installed the edge 5.19.4 kernel
@fabian In case there's an issue with this kernel, is there any kind of deep debugging we can do to try and nail down this issue? I've set up netconsole & kernel debugging from within the VM from this page: https://wiki.ubuntu.com/Kernel/Netconsole . Is there anything else we can do to capture this panic and provide as much information for you?
Code:
root@pve:~# uname -a
Linux pve 5.19.4-edge #1 SMP PREEMPT_DYNAMIC PVE Edge 5.19.4-1 (2022-08-25) x86_64 GNU/Linux@fabian In case there's an issue with this kernel, is there any kind of deep debugging we can do to try and nail down this issue? I've set up netconsole & kernel debugging from within the VM from this page: https://wiki.ubuntu.com/Kernel/Netconsole . Is there anything else we can do to capture this panic and provide as much information for you?
I was using the same kernel as yours, but my old CentOS 7 vm failed (crashes every few hours.)
I'm now on Fabian's 5.19 version and I'll report back.
My 2 VMs that crashed/rebooted every 1-2 hours are a Cent7 and Fedora 36 both running newest epel kernel-5. Doesn't appear to be the kernels inside the VMs but the proxmox kernel.. Going strong now 14 hours. We'll see in a day or 2 if it's stable long term.
For my part, I have 2 VMs out of the 3 that hold up. I have my "IoT" VM which went wrong 2x with an uptime of +- 5-6h.
Here, I relaunched it yesterday and it seems to hold.
The "System" VM is holding up, it's been running for as long as the host... So even if it's not 100% reliable yet in 5.15.39-4, I remain positive because there is an improvement.
I will also try the kernel edge and give my feelings.
@thimplicity try disabling the "KVM Hardware Virtualization" by going to VM->Options
Cold boot the VM
Cold boot the VM
2 days and 9 hours of uptime since I installed the 5.19.4 edge kernel... rock solid.