Re: Can't start VM after latest kernel update: weird mount failure

I can get around the mount problem by explicitly specifying the filesystem type (e.g., mount -t ext2 /dev/vda1 somedir)--is that some limitation of busybox mount?

I still don't know why the system is having trouble starting up.

The recent kernel update included a fix for a KVM vunerability(https://security-tracker.debian.org/tracker/CVE-2015-5307). Perhaps that has something to do with it.

Is there a way I can proceed by hand after I mount the proper file systems?

Ross

P.S. I also tried adding noresume to the options on grub's linux line; it didn't help.

On Thu, Nov 12, 2015 at 1:57 PM, Ross Boylan <rossboylan@stanfordalumni.org> wrote:

After the recent kernel updates one of my virtual machines won't start. GRUB runs and messages indicate the kernel is loaded and the initial ramdisk is loading. Then it says it can't find the root device (identified by correct UUID, though it wouldn't be visible until the logical volumes are activated). In busybox the logical volumes are visible, but when I make a directory and attempt to mount to it I get
mount: mounting /dev/markov02/root on r failed: No such file or directory.
The same thing happens if I try to mount the boot partition (/dev/vda1).
This is weird because both the device and the directory are present. I can change into the directory and create a file in it

When I attach the virtual hard drives to another VM I have no problem mounting either the boot partition or the logical volume.

I would appreciate any assistance figuring out what's going on or what I can do to correct it.

If my initrd got corrupted, is there a way to regenerate it?

Thanks.
Ross Boylan