issue with fstab and fsck after upgrade

[Matt Morgan]

On Nov 21, 2007 7:00 PM, Chris Dawson <xrdawson at gmail.com> wrote:
> I think this is an issue with my upgrade to Gutsy, but I am not sure.
>
> Whenever I boot the system drops me into a subshell and then tells me
> I need to fsck.  I've booted using the live CD, then installed lvm2
> and run the series of commands to mount as an lvm partition, and then
> run fsck and there were no errors.  If I enter "halt" the subshell
> exits and things are generally normal, except that I always have to
> run "sudo hostname arrail" to reset the hostname, then log out and log
> back in.
>
> I suspect this is an issue with my /etc/fstab file as it does not look
> the same as other machines I have that also have lvm partitions.  One
> note:  I had /dev/sda2 mapped to a USB disk (ipod, commented out
> below), and I wonder if that confused the upgrade system as the
> machine thinks my /dev/hda is /dev/sda, and sda2 is my main disk
> partition.  Perhaps it wanted to create a UUID mapping, but saw a
> conflict between those disks and just aborted leaving me with a weird
> fstab?
>
> This is my fstab:
>
> cdawson at arrail:~$ cat /etc/fstab
> # /etc/fstab: static file system information.
> #
> # <file system> <mount point>   <type>  <options>       <dump>  <pass>
> proc            /proc           proc    defaults        0       0
> /dev/mapper/vg-root /               ext3    defaults,errors=remount-ro 0       1
> # /dev/hda1 -- converted during upgrade to edgy
> UUID=fbc170cc-bf5e-4aa6-b77a-a3d129715077 /boot ext3 defaults 0 2
> /dev/mapper/vg-swap_1 none            swap    sw              0       0
> /dev/hdc        /media/cdrom0   udf,iso9660 user,noauto     0       0
> /dev/fd0        /media/floppy0  auto    rw,user,noauto  0       0
> #/dev/sda2        /media/ipod   auto    rw,user,noauto  0       0
> 192.168.2.8:/mnt/windows-box-fs /mnt/windows-box-fs nfs noauto 0 0
>
> Should it not be that I have two drives with UUID's?  How does the
> system know that something is corrupted on boot if an lvm partition is
> not recognized at that point in the boot process?
>
> Here is what I get when I run df:
>
> cdawson at arrail:~$ df
> Filesystem           1K-blocks      Used Available Use% Mounted on
> /dev/mapper/vg-root   75154576  63768800   7568100  90% /
> varrun                  257768        84    257684   1% /var/run
> varlock                 257768         4    257764   1% /var/lock
> udev                    257768       404    257364   1% /dev
> devshm                  257768    134884    122884  53% /dev/shm
> lrm                     257768     35324    222444  14%
> /lib/modules/2.6.22-14-386/volatile
> sshfs#root at localhost:/
>                      1048576000         0 1048576000   0% /home/cdawson/piab
> /dev/mapper/sda1        233335     84832    136055  39% /boot
>
>
> And, here is what I get when I run lvdisplay:
>
> cdawson at arrail:~$ sudo lvdisplay
>   --- Logical volume ---
>   LV Name                /dev/vg/root
>   VG Name                vg
>   LV UUID                kPLR83-a13z-6gqG-te29-qTgt-QRNL-CIC4b9
>   LV Write Access        read/write
>   LV Status              available
>   # open                 1
>   LV Size                72.82 GB
>   Current LE             18641
>   Segments               1
>   Allocation             inherit
>   Read ahead sectors     0
>   Block device           254:4
>
>   --- Logical volume ---
>   LV Name                /dev/vg/swap_1
>   VG Name                vg
>   LV UUID                6I3jCL-lsfO-pjH9-Dgnd-aPhn-2vyQ-7zXFie
>   LV Write Access        read/write
>   LV Status              available
>   # open                 1
>   LV Size                1.47 GB
>   Current LE             377
>   Segments               1
>   Allocation             inherit
>   Read ahead sectors     0
>   Block device           254:5
>
> --
> ubuntu-users mailing list
> ubuntu-users at lists.ubuntu.com
> Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users
>

Roughly the same thing happened to me when I upgraded to Gutsy, and I
see at least one other message on the list, from Robert Fitzpatrick,
with a similar problem.

In my case, this is fstab after I upgraded:

proc            /proc           proc    defaults        0       0
# /dev/hdc3
UUID=562997ca-faa8-45bf-aca9-975ce010bb0e /               ext3
defaults,errors=remount-ro 0       1
# /dev/hdc1
UUID=C8C82629C826166A /media/hdc1     ntfs
defaults,nls=utf8,umask=007,gid=46 0       1
# /dev/sda1
UUID=ee7a0a32-cf2d-4300-a4f5-52e24c8f62df /opt     ext3    defaults
    0       2
# /dev/hdc2
UUID=21bf1e1a-ce19-4354-a4bf-e56539a0e8c6 none            swap    sw
           0       0
/dev/hda        /media/cdrom0   udf,iso9660 user,noauto     0       0
/dev/fd0        /media/floppy0  auto    rw,user,noauto  0       0
/dev/sda2 /home     ext3    defaults        0       2

When I booted with that fstab, my /home and /opt would not be mounted,
since they're actually on /dev/sdb, which isn't even mentioned here
(and I guess the UUIDs are wrong). I'd get dumped to single-user mode
with the request to fsck; I didn't really have to, though, I just had
to mount home and then 'exit.'

I changed fstab to

proc            /proc           proc    defaults        0       0
# /dev/hdc3
UUID=562997ca-faa8-45bf-aca9-975ce010bb0e /               ext3
defaults,errors=remount-ro 0       1
# /dev/hdc1
UUID=C8C82629C826166A /media/hdc1     ntfs
defaults,nls=utf8,umask=007,gid=46 0       1
# /dev/hdc2
UUID=21bf1e1a-ce19-4354-a4bf-e56539a0e8c6 none            swap    sw
           0       0
/dev/sdb1 /opt     ext3    defaults        0       2
/dev/sdb2 /home     ext3    defaults        0       2
/dev/hda        /media/cdrom0   udf,iso9660 user,noauto     0       0
/dev/fd0        /media/floppy0  auto    rw,user,noauto  0       0

and now it works fine. Basically it looks like there are some
conditions in which the gutsy upgrade gets fstab wrong.