PANIC: double fault, error_code 0x0 - IBM x3650 M3

[czbg]

Hi,

every week or so, this one server crash with PANIC: double fault, error_code 0x0. I've made some screenshots(attached) of console in that state. I can only reboot after that. Anyone have any idea? I've tun some onboard IBM diagnostics, Linux Stresstest, memtest, all pass w/o errors.

Proxmox is latest from unsupported repo. This is only IBM that I have here and It's the only server with this problem.

Thank you and Best Regards

 

[talos]

Hi,

i have the same issues with two x3650 M3. One of the machines freezes every week and light path reports memory fault. replaced memory, same issues and no errors with extensive memtest checks.

Second machine just randomly reboots.

These are spare machines with no active vm/lxc on it.

 

[czbg]

Hi talos,

thanks for that info. This server also has only a few ct/vm, for testing. I do not know what to do, I cannot return it and claim warranty because they will just run onboard diagnostics and say all is ok.

Problem happens with different kernels. Since 4.x release, I've upgraded it a few times from unofficial repo, every time there was a new kernel and I hoped it would change but it didn't. From pve-kernel-4.2.3-2-pve to pve-kernel-4.2.8-1-pve

 

[talos]

Yeah i have no idea what to do. This machine is just for spare / testing stuff. i should take a look at ubuntu with lxcd.