Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 09 Jan 2014 03:48:06 +0100
From:      Julian Stecklina <jsteckli@os.inf.tu-dresden.de>
To:        kpneal@pobox.com
Cc:        Thomas Hoffmann <trh411@gmail.com>, FreeBSD Stable <freebsd-stable@freebsd.org>
Subject:   Re:  UFS panic on Google Compute Engine
Message-ID:  <1389235686.3057.4.camel@janus.xn--pl-wia.net>
In-Reply-To: <20140108193938.GA19267@neutralgood.org>
References:  <lajo13$2g6$1@ger.gmane.org> <CAB7-odkZAGxgtfnqNwNaCfqXcY0nmbFPdf3NtFYkwGUictJcyw@mail.gmail.com> <20140108193938.GA19267@neutralgood.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mi, 2014-01-08 at 14:39 -0500, kpneal@pobox.com wrote:
> On Wed, Jan 08, 2014 at 01:05:55PM -0500, Thomas Hoffmann wrote:
> > On Wed, Jan 8, 2014 at 9:40 AM, Julian Stecklina <
> > jsteckli@os.inf.tu-dresden.de> wrote:
> > 
> > > Hello,
> > >
> > > I have set up a FreeBSD 10.0 RC4 system on GCE using this [1]
> > > description. I have hit the following panic on two different setups when
> > > trying to do portsnap fetch extract and using portinstall:
> > >
> > > panic: ufs_dirbad: /: bad dir ino 24563 at offset 0: mangled entry
>  
> > I experienced that exact same panic back on 9.1-RELEASE. Every time I
> > executed a command that used /var, my system would panic. I cleared it by
> > running fsck -y on all filesystems while in single user mode, so that is
> > the first thing I would try in spite of the fact that (unfortunately)
> > running fsck on GCE is a "bit hard". Sorry.
> 
> The easiest thing to do might be to hack /etc/rc.d/fsck to do the
> 'fsck -y' unconditionally. Change the '-p' to '-y' in /etc/rc.d/fsck
> and then reboot. After the problem is solved be sure to change it
> back to the way it was.

Thanks for the tip. Too easy. ;) The filesystem is seriously damaged:

Starting file system checks:
/dev/da0p2: INCORRECT BLOCK COUNT I=164645 (8 should be 0) (CORRECTED)
/dev/da0p2: INCORRECT BLOCK COUNT I=164650 (8 should be 0) (CORRECTED)
/dev/da0p2: INCORRECT BLOCK COUNT I=164654 (8 should be 0) (CORRECTED)
/dev/da0p2: INCORRECT BLOCK COUNT I=164659 (8 should be 0) (CORRECTED)
[... lots of those ...]
/dev/da0p2: UNALLOCATED  I=4378  OWNER=root MODE=0
/dev/da0p2: SIZE=0 MTIME=Jan  1 00:00 1970 
/dev/da0p2: NAME=/usr/ports/math/lapack/files

/dev/da0p2: UNEXPECTED SOFT UPDATE INCONSISTENCY; RUN fsck MANUALLY.
Automatic file system check failed; help!

I checked the local copy of the image I uploaded to GCE and its
filesystem is fine. Must be happened during transfer to GCE. Weird.

Sorry for the noise,
Julian





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1389235686.3057.4.camel>