Re: root filesystem corrupted, can't log in
Bob,
np.
I intended to use -r. Below your post is a response to the same question
from another list member. All I can say is it seems to work for me. Next
time I'll try it again without the -r. I'm sure I'll have occasion to.
montefin
"Robert D. Hilliard" wrote:
>
> montefin <montefin@finux.com> writes:
> > NOTE: If you get a large block of type that mentions a SUPERBLOCK error
> > and tells you to do the '-b 8193' thing. Well, that's how you got that
> > message, right? Instead I did this:
> >
> > e2fsck -r /dev/<filesystemname> (my /var looks like /dev/hda7
>
> With e2fsprogs 1.18-3, the e2fsck manpage says:
>
> -r This option does nothing at all; it is provided
> only for backwards compatibility.
>
> What option did you intend to use? (I'm not flaming - I'm
> asking for information.)
>
> Bob
> --
Yeah, I know, I read that too.
But I'd seen it mentioned in one of Barkakati's books, I think, so I
tried it and it worked for me when I was in the same jam as you. And it
didn't work without it. But maybe that's just Red Hat or my environemnt,
although it's worked for other people I passed it along to. -r is a hold
over from FSCK and it's supposed to be e2fsck's default behavior, but
there's a lot that isn't quite so about e2fsck.
Maybe it's just voodoo for me now, but, hey, if it doesn't do anything,
it can't hurt, right?
Read those other comments about the internal directives in the e2fsck
program. The -b 8193 alternate superblock routine is suggested after
_it_ produces the error. And wait until you see the 'bonehead' line in
the "maybe it's a bug but it's probably YOU" message. Anyway, I tend not
to believe everything I read in man pages.
Seriously though. Check to see if you're running e2fsprog along with or
as a part of e2fsck on your bootup. It's a little script that shows the
pretty hash marks or spinning pinwheel for e2fsck's progress. Well, it
was also known to trigger an unnessecary Signal 11 everytime the root
filesystem was checked. Not only for a power outage but even when you
reached maximum boots without a check. That's no fun.
There's a bugfix on the Red Hat 6.1 Errata page
http://www.redhat.com/support/errata/rh61-errata-general.html
that supposedly corrects that, but I never saw it make a difference. I
think it's just e2fsck. I'm not sure if Red Hat fixes can be applied to
other Linuxes, but you might want to try it.
Reply to: