fsck doesn't like LABEL= in fstab
Giles Orr
gilesorr-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org
Tue Oct 9 01:28:56 UTC 2007
What with my hard drive adventures recently, I decided it was time to
muck with my /etc/fstab some so that USB devices would start showing
up at coherent mount points. I thought the right method would be
using "UUID=..." statements in fstab, but then I found "LABEL=..."
which seemed more practical. However, I've had the same issue with
both: on boot, fsck tries to check the UUID or LABEL devices (which
are generally not connected), despite "noauto" params, and then stalls
the boot process suggesting I go fix it. Here's the
/var/log/fsck/checkfs :
Log of fsck -C -V -R -A -a
Mon Oct 8 17:04:22 2007
fsck 1.40.2 (12-Jul-2007)
Checking all file systems.
[/sbin/fsck.ext3 (1) -- /fc] fsck.ext3 -a -C0 /dev/hda1
/: clean, 178437/1310720 files, 932982/2620595 blocks
[/sbin/fsck.ext3 (1) -- /mnt/vantec1] fsck.ext3 -a -C0 LABEL=Vantec01
fsck.ext3: Unable to resolve 'LABEL=Vantec01'
[/sbin/fsck.ext3 (1) -- /mnt/vantec2] fsck.ext3 -a -C0 LABEL=Vantec02
fsck.ext3: Unable to resolve 'LABEL=Vantec02'
fsck died with exit status 8
The corresponding lines in /etc/fstab :
/dev/hda1 /fc ext3 defaults,noauto 0 2
LABEL=Vantec01 /mnt/vantec1 ext3 user,noauto 0 2
LABEL=Vantec02 /mnt/vantec2 ext3 user,noauto 0 2
So what's the problem here? If I use "/dev/sda1" instead of
"LABEL=Vantec01" fsck ignores the statement because it says "noauto"
(see below) but change it to "LABEL=..." or "UUID=..." and I have a
fairly nasty issue. Although writing this did make me notice that /fc
is "noauto" and gets checked. But I have other lines:
/dev/sda1 /mnt/usba auto user,noauto,noatime 0 0
/dev/sdb1 /mnt/usbb auto user,noauto,noatime 0 0
Which are ignored. I'm a bit confused.
--
Giles
http://www.gilesorr.com/
gilesorr-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org
--
The Toronto Linux Users Group. Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists
More information about the Legacy
mailing list