Boot error because e2fsck thinks ext3 filesystem is ext2?


During the file system check of an ext3 partition at boot I get the following output:


[successful fsck output]

fsck.ext3: Bad magic number in super-block while trying to open /dev/sdc6
The super-block could not be read or does not describe a correct ext2 filesystem. If the device is still valid and it really contains an ext2 filesystem (and not swap or ufs or something else), then the superblock is corrupt, and you might try running e2fsck with an alternate super-block:
e2fsck -b 8193 <device>

[more successful fsck output]

fsck failed for at least one filesystem (not /).
Please repair manually and reboot.
The root filesystem is already mounted read-write

I’m then forced to login in as root and given the following prompt:

(repair filesystem) #

I’m reluctant to do as advised by the output and run e2fsck -b because it is not an ext2 filesystem.

Although I can still enter runlevel 5, it doesn’t seem to recognise mouse and keyboard input in KDE so my system is effectively redundant at the mo. For this reason any short term workarounds are welcome, but a fix is needed. This problem is part of a longer saga to do with recovering a Windows Vista installation which started failing to boot. I have used both Vista and SUSE tools to try and recover my bootloader to no avail, and this has been the result. If more detail about this is needed please ask and I can explain what I have done.

Any help would be greatly appreciated

Many thanks


is /dev/sdc6 vital for the operation of your computer? if not remove it from /etc/fstab or comment it out #
is it a usb disk drive? if it is make sure its attached to your computer
e2fsck is valid for ext2,3,4

No, it’s not an essential partition. I followed your advice and I’m up and running again. Thank you very much for your prompt attention!

I’m a bit puzzled as to why it doesn’t like that partition all of a sudden though. I’ve had no trouble with it being in fstab before and I would prefer that it was mounted at boot time if possible. Is it worth trying to correct the superblock do you think?

There must be something siriously wrong on that fs. The fact that it talks about ext2/ext3 indiscriminatly does not mean it sees it wrong. Both are much the same, only the journaling is the difference, but the main lay-out is the same. Thus when it says “Bad magic number in super-block”. I should take that serious.

Adding another drive, even a SD card can through the sequence
off and sdc can become sdb when the kernel tries to mount sdc
it tries to run a file check against a partition that it can’t see
and returns the bad blocks error message
A while back I tried to put my home partition in an external drive and kept running into the same problem.

Does any body know how to turn off the automatic fsck on a partition??

After rereading this thread the culprit is /dev/sdc6 so while you are logged in as root do a fsck -f -r on that partition
also read man fsck

You should not have your partitions mounted by /dev/sda1, etc (in /etc/fstab). but by /dev/disk/by-id/… as openSUSE does by default since several levels. Thus the mounting will be persistent even if the sda/sdb/sdc/sdd is not.

It is not a good idea to stop all sorts of good health checks, like it is not a good idea to drive without seatbelts or to connect to the internet without firewall. The OP now got a warning that something is wrong and he can try to save as much as he can (maebe even repair). What do you think that would have happened when it simply had mounted a broken fs?

Do a ‘man fstab’. The last 2 digits on a fstab entry are about file system checking on boot.

To the OP: I agree with Henk. Serious trouble. If I see things like this, I replace the disk immediately. Then I dd the new disk to the old one, usually doesn’t work, just to test if it’s indeed dead or in a ‘possibly degrading condition’.

Trouble would be the result
Isn’t fsck available in the repairfs I think a coupla times I had o run the fsck to get my system up and running. Found out what the lost+found folder was for

Yes, hurray for lost+found. But that is for the ‘easy’ fs problems. When the magic number is wrong, it is serious.

The fact it says there’s bad magic in the superblock means something went really wrong with your file system. However, Ext2/3/4 has a backup superblock which can be restored in case the primary one got corrupted or wiped out

Assuming your file system uses 4K blocks (the default), then run fsck as follows:

e2fsck -b 32768 /dev/sdc6

(the 32768 is the block location of the backup superblock in a 4K-per-block Ext file system. It is dependent on the block size of the file system so if you use another one (1K, 2K, etc) it’ll be at another block location so consult the man page)

And no, e2fsck does not think your file system is Ext2. It shows Ext2 in its warning text because both Ext3 and Ext2 are essentially the same, except the former has a journal