You are not logged in.
Hello
My Arch Linux based home server seems to have a problem with virtualization running libvirtd and qemu.
The hardware that I have in the computer is as follows:
CPU: Intel Core 2 Quad Q9450
Motherboard: ABIT IX38 QuadGT (Intel VT-d enabled in the BIOS and passes checks for it in the OS too.)
RAM: 8Gb
GPU: ATi Radeon HD 4670 (Mainly just there to give me a graphical console when I hook up a monitor.)
Harddrives: 4x 1.5Tb Seagates, 2x 2Tb Seagates (These 6 drives of different size, both host a mdadm RAID 6 array.) and a 160Gb WD IDE drive as the drive that hosts the OS, not the virtual machines.
On this "host" I got 4x VMs running via Qemu:
1x Windows Server 2008 R2
2x Ubuntu 12.04 Server
1x PFSense FreeBSD 10.1
Since I built this machine way back in 2010 and ran Ubuntu 10.04 LTS on it, I decided that this year after support ended for 10.04 LTS, to move on to a more updated distro.
For this I chose Arch Linux, and with it came a few changes, rather than VMWare Workstation I had to transition the VMs over to Qemu, which I did with drive cloning and a little manual labor in the form of copying settings manually, and it seemed to work fine.
However, after a couple of days or sometimes even hours, I get a kernel panic.
The kernel panic have been eluding all efforts to capture it in the logs up until today when I finally managed to capture its seemingly never ending stack trace.
When the kernel panic occurs, I don't know if it starts in the guest or not, but everything on the host dies with it and I can only reboot as a means to recover.
If this is a bug in the kernel as the log seems to state it is, then consider this thread as a means to hash out all the needed information to file this to the bug tracker.
Attached is a 90ish kb text file of the stack trace for this kernel panic that doesn't make much sense to me other than point to qemu: https://www.mediafire.com/?7a7e78u6oqs8iqy
Thank you for the time anyway.
Offline