My system crashed while trying to do a dump to the tape drive. I get
a "bad super block magic number wrong" message that fsck can't fix.
Is there a way to fix this without killing all my files?
The superblock# are now printed. You can subsequently run
fsck -b # where # is one of the block# printed by newfs -N
Hope this helps
There's an fsdb around (it went to /xperimnt in 2.1R and is in the
base system in the 2.2 development sources), but it's perhaps rather
tricky to use it. Try if you can use alternate superblocks (fsck -b
Never trust an operating system you don't have sources for. ;-)
Re-installed 4.8 after formatting the disk. Just as a test, went down
to single-user mode, un-mounted all the file systems, and ran fsck
which reported no problems. For curiosity, I ran fsck -p /dev/ad0. It
reported that there was a "bad super block: magic number wrong" and
unexpected inconsistency: run fsck manually.
This message was ignored and the system was rebooted with no problems.
A new kernel was built and no problems occurred. It seems that the
file system is working perfectly well, in spite of this "bad super
The question is, is it normal to have some trivial errors be reported
by fsck at any given invocation? Is there nothing wrong with my file