Bug#612809: fsck failing because it can't find fsck.btrfs
Jonathan Nieder wrote: > Still, I believe it would be best to allow btrfsck to act as fsck when > the filesystem is not broken. Would some script along these lines > make sense? Alas, the answer appears to be "no" for now. warning: btrfsck cannot repair filesystem corruption - suppressing "-a" option check_mounted(): Could not open /run/rootdev Could not check mount status: Unknown error 18446744073709551610 fsck died with exit status 250 Thanks, and sorry for the noise. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#612809: fsck failing because it can't find fsck.btrfs
I have same trouble too. My work around was working on other systemthat: cp /path/to/root/sbin/btrfsck /path/to/root/sbin/fsck.btrfs But, it hasn't been a good hack. As you know, btrfsck's synopsis is very simple: btrfsck device But fsck.hoge did -a option. So, while boot, an error saying "-a device is not available" or so. And btrfsck does not do anything but just read. Judging from above, this way is possibly to be a good hack, ln -s /bin/true /sbin/btrfsck and/or Mail to root when a system dist-upgraded to recommend that: In fstab, the last column of / would be 0 Anyway, it is not friendly way when dist-upgrade, turned off, and reboot, then there is no rule to fsck.btrfs. Thanks.
Bug#612809: fsck failing because it can't find fsck.btrfs
found 612809 btrfs-tools/0.19+20101101-1
quit
Hi,
Alan Chandler wrote:
> This bug seems to have got a whole lot worse by the last upgrade to my
> system because now I am thrown into maintenance mode with root mounted
> read-only.
Ran into this today (in a VM I hadn't upgraded for a while); looks
like it was caused by
* Removing fsck.btrfs symlink to circumvent that btrfsck doesn't
support -a.
in version 0.19-11. Which makes sense --- btrfsck doesn't know how to
repair filesystems with guidance from the user yet, so it certainly
can't automatically repair them.
Still, I believe it would be best to allow btrfsck to act as fsck when
the filesystem is not broken. Would some script along these lines
make sense?
#!/bin/sh
# fsck.btrfs - Wrapper for btrfsck to work around the latter's
# lack of support for "-a".
shell_quote () {
# Wrap stdin in single-quotes,
# quoting embedded single-quotes as '\''.
sed -e "s/'/'''/g" \
-e "s/.*/'&'/"
}
args=
while test "${1+set}"
do
if test "$1" = -a
then
echo >&2 'warning: btrfsck cannot repair filesystem
corruption!'
else
args="$args $(printf "%s\n" "$1" | shell_quote)"
fi
shift
done
eval "exec btrfsck $args"
Thanks,
Jonathan
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#612809: fsck failing because it can't find fsck.btrfs
This bug seems to have got a whole lot worse by the last upgrade to my system because now I am thrown into maintenance mode with root mounted read-only. My solution was to provide a symbolic link so that /sbin/fsck.btrfs points at /sbin/btrfsck I still seem to get some form of failure reported during boot up, but at least it does finish at put me into the gdm prompt -- Alan Chandler http://www.chandlerfamily.org.uk -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
