Actually I don't really understand the problem or solution. Could you please elucidate in case someone else runs into this?
I had run Parted Magic LiveUSB. I did this to clean up some old unused partitions with GParted and grow the existing /mnt/DATA and /var into the freed up space. These were all logical partitions. Then I removed some /etc/fstab entries for a HDD which was failing and I had decided to remove and not replace (It was just a Windows XP install for school work and a large partition with movie files which I never watch).
Then I decided that I should create images of my new set up. No, I did not reboot at this point, so the problem did not raise it's ugly self.
After running Clonezilla twice, once for an image of my SSD (/, /boot, /home) and once for the HDD (swap, /var, /tmp, /mnt/DATA) I rebooted the machine and had the problem appear.
This error happens after the Grub splash screen when the filesystems are being checked and mounted, before the login screen.
It says, Bad magic number detected in the super-block /dev/sdb8. And a bunch of stuff about is it a swap or ext filesystem. Proposed fix on screen is to run
e2fsck -b 8130 /dve/sdb8
to fix it. Then it tells you to enter the root password to perform maintanence or CNTL+D to continue (which fails and reboots to same error).
I entered root password, ran the proposed solution and received the same error message as it's output. Read the man page for e2fsck and found that if your filesystem uses different block sizes you need to replace the 8130 (which is for 1k block) with another number which is provided in the man page. I tried that and had the same output. I then ran fdisk -lu to see if there was anything really obvious missing. I had just re-arranged a bunch of partitions on that disk and removed another completely. Nothing jumped out at me right then.
After that I Googled a bit and got a bunch of really complex, technical fixes... none sounded as if they would actually fix the problem and they could cause complete loss of the filesystem. Then I posted here and in the process of typing the post I realized that /dev/sdb8 was listed in my /etc/fstab as /tmp and fdisk as swap. That got me wondering if there was an easy fix. It seemed like removing and resizing some partitions allowed them to be renumbered and fstab entries were no longer accurarte. the OS uses /etc/fstab to check and mount the filesystems at boot.
I rebooted, entered the root password for maintanence, remounted the root filesystem read/write, and edited my /etc/fstab to reflect the values I found with
And viola, the machine booted with no errors.
Great job Amenditman
Ahhh ... I'm completely lost on the issue in this thread. As it appears, your mount points in /etc/fstab were wrong? Or got edited by Clonezilla? It must have been running OK before???
And you use separate partitions for /tmp and /var? Your choice, or done by an install program, or ???
I have seen complaints like that on my own machines, but always due to my own editing errors when I go to add stuff to fstab.
It was running fine before I began this weekend project.
I use separate partitions for /var and /tmp by my own choice.
Did I miss something here?
Nope, you can go back to sleep.