On January 5, Eric Sandeen wrote:
> On Mon, 5 Jan 2004, Rainer Krienke wrote:
> > For one filesystem (~150GB) this worked. xfs_repair reported some errors in
> > the filesystem but finished its work. Next I tired to mount this filesystem
> > but mount complained that it could not find a valid superblock. So I ran
> > xfs_repair once again. It still found some errors (but less than before).
> > Next I rebootet the machine and the filesystem was mounted.
> Unable to find a superblock immediately after repair? I have never
> seen this before, sounds very odd. BTW running repair twice in
If this still happens, It might be useful to see how much was
overwritten, and if you recognise the data. You can use xfs_db to
read and print the superblock.
xfs_db -r /dev/diskxyz
print - a formatted superblock.
print - a hex dump of the superblock
> > - Is it a known issue that xfs_repair seg faults sometimes or is it perhaps
> > a
> > problem of my version (see below) ?
> I wonder if maybe it's failing a memory allocation for the large
> filesystem; I thought i remembered a (fixed) problem like this
> but I don't see it in the changelogs. A little gdb debugging
> would be a big help.
xfs_repair can "panic" in which case it calls abort() so you get a
coredump to look at. It's a bit unfriendly in that it doesn't exactly
say it's going to do that . . .