How to perform an FSCK
Posted by mark1hos, 09-16-2007, 02:52 PM Hi guys, we've got our server hooked up to a KVM, as the server has experienced a kernel panic, however, I was wondering how we can force an FSCK?
Posted by Abee, 09-16-2007, 03:10 PM fsck should always be run in a single user mode which ensures proper repair of file system . If it is run in a busy system where the file system is changing constantly fsck may see the changes as inconsistencies and may corrupt the file system . For further details please take a look at http://www.adminschoice.com/docs/fsck.htm
Posted by mark1hos, 09-16-2007, 03:11 PM Hi, thanks for this. So can the server perform an FSCK with a kernel panic? Because our datacentre is saying it cannot and it needs a new hard drive and O/S
Posted by spulis, 09-16-2007, 03:49 PM You need to first find the cause of the panic. Was the kernel changed? You should be looking through your logs to see what caused the panic.
Posted by mark1hos, 09-16-2007, 04:27 PM Yes we would but it does not get passed the kernel panic.
Posted by spulis, 09-16-2007, 04:30 PM So has anything on the system changed? Do you have a remote console on the box to see at what point it panics?
Posted by mark1hos, 09-16-2007, 04:37 PM Nothing at all has changed. It was just a routine monthly reboot and it just did'nt come back up. It loads a blue fedora page which loads and then lots of text and gets to umount /initred/dev/ failed: 2 Kernel panic - not syncing: Attemped to kill init! [ ] panic+0x42/0x1ca
Posted by spulis, 09-16-2007, 04:56 PM You can try booting from a rescue disk and perform a fsck on the root disk and check the disk utilization on the drive. Is tmp full or another volume maxed out. Does it mention anything about a missing disk label?
Posted by mark1hos, 09-16-2007, 04:57 PM I do not know it says a lot of stuff none of it I actually understand. I am not sure my datacentre will use a rescue disk. They said they are going to put a new harddrive in tomorrow morning.
Posted by mark1hos, 09-16-2007, 05:53 PM Please find attached the error message. Attached Thumbnails
Posted by derek.bodner, 09-16-2007, 09:50 PM don't worry about running an fsck until you get a bootable kernel. Sounds to me like either: - The SATA/SCSI controller isn't compiled into the kernel correctly - The MBR's corrupt. I'd have them boot it into a live CD, startup ssh, set a root password, ssh in, chroot to the previous environment, and re-setup grub and double check the kernel (and recompile if necessary). Last time you ran updates (yum, up2date, apt, emerge, whatever), were any of the updates kernel related?