CentOS + Bad Plesk Update = Kernel Panic

After an unscheduled reboot (breaker went out) we had a CentOS web server running Plesk coming up with a kernel panic on boot – actual error:

“Mounting root filesystem
mount : error 19 mounting ext3,usrquota
mount : error 2 mounting none
Switching to new root
Switchroot : mount failed : 22
umount /initrd/dev failed : 2
Kernel panic — not syncing … ”

After some unneccessary fscking and fudling around I found this post mentioning that this was potentially the result of a Plesk 8.1 bug which generates a bad init image as part of a system update. The post narrows it down to a badly configured fstab entry by Plesk which is then rolled into the new init, but what was really important for me was simply that the init image was what was causing the problem. Surely enough, when I selected the alternate kernel in grub during boot the system came up without a hitch. For the record, my excuse for not trying the old kernel boot line in the first place is that this whole deal was going down at 4:30 A.M. and we were in a bit of a rush, among other things.

Long night, happy ending with a quick fix. Now my Sunday is screwed…

Leave a Reply

Your email address will not be published. Required fields are marked *