On 2 different VMs, the installation of vmware-tools has caused my SLES guests to go into a kernel panic after reboot. It was the same version of VMwareTools in both cases, so I'm thinking it's not a coincidence. Also, both VMs were running SLES9 and using the lilo boot loader. hmm...
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(8,2)
Server1 - SLES-9-i386-SP4
server1:~ # rpm -qa | grep -i vmware
VMwareTools-8194-208167
Server2 - OES-9-i386-SP2 (based on SLES9)
server2:~ # rpm -qa | grep -i vmware
VMwareTools-8194-208167
Fortunately, I found a solution here.
I booted from a rescue CD, rather than the SLES install disk. I used SystemRescue CD, but I'm sure just about any live cd distro would work. Once I booted from the iso file, I used gparted to see the partitions on my disk. I had a swap partition at /dev/sda1 and the file system was at /dev/sda2. The troubleshooting reference above says to mount your partition at /mnt. I'm not sure that's safe so I mounted /dev/sda2 at /mnt/custom instead.
systemrescue:~ # mount /dev/sda2 /mnt/custom
systemrescue:~ # mount -t proc none /mnt/custom/proc
systemrescue:~ # mount -o bind /dev /mnt/custom/dev
systemrescue:~ # cd /mnt/custom
systemresuce:~ # chroot /mnt/custom
root:~ # mkinitrd
root:~ # lilo
There were a couple of warnings that popped up - specifically the dependencies for the vmxnet module couldn't be resolved. This has to be part of the problem in the 1st place. After mkinitrd finished, it prompted me to run "lilo" from the command line. Then I rebooted and the server came up clean.
Comments