[Bug 994749] Re: checkfs run on already mounted filesystem, preventing boot

[GB]

I'm not sure why you flagged this as 'util-linux"

I think the culprit may be ureadahead

Please see the attached bootchart.

ureadahead calls mountall before 30checkfs.sh runs.

Why?


** Attachment added: "theta-precise-20120505-1.png"
   https://bugs.launchpad.net/ubuntu/+source/util-linux/+bug/994749/+attachment/3132695/+files/theta-precise-20120505-1.png

-- 
You received this bug notification because you are a member of Ubuntu
Foundations Bugs, which is subscribed to util-linux in Ubuntu.
https://bugs.launchpad.net/bugs/994749

Title:
  checkfs run on already mounted filesystem, preventing boot

Status in “util-linux” package in Ubuntu:
  New

Bug description:
  I had Ubuntu 10.04 LTS installed, now upgraded to 12.04 Precise.

  When the system enters run state 'S', it attempts to fsck two mounted
  file systems. This causes it to drop into a maintenance shell,
  preventing other system services (sshd, etc. etc. ) from starting
  until the maintenance is entered.

  However, it does progress to running gdm and gives a prompt to login
  (X is running on tty1 then).

  Ctrl-Alt-F7 leads to the maintenance shell, exiting it makes the boot
  complete.

  /var/log/checkfs is
  ----------------------
  Log of fsck -C -R -A -a 
  Fri May  4 15:40:13 2012

  fsck from util-linux 2.19.1
  /dev/sda2 is mounted.  e2fsck: Cannot continue, aborting.

  /dev/sda4 is mounted.  e2fsck: Cannot continue, aborting.

  fsck died with exit status 8

  Fri May  4 15:40:13 2012
  ----------------

  My /etc/fstab:

  proc            /proc           proc    nodev,noexec,nosuid 0       0
  /dev/sda1       /               ext4    errors=remount-ro 0       1
  /dev/sda2       /home           ext4    defaults        0       2
  UUID=02e8ad70-9e55-41e3-85be-42d7edf07c5e /opt            ext4    defaults        0       2
  /dev/sda3       none            swap    sw              0       0
  /dev/sdb3       /media/gback    ext4    rw              0       0

  (The device with UUID=02e... is /dev/sda4).

  My question: how in the world can it happen that your checkfs.sh runs
  'fsck -C -R -A -a' *apparently after* you've mounted all configured
  filesystems read-write?  This would fail deterministically, as it does
  for me. Clearly, something must be wrong with my installation, but
  what? I did not make any changes to /etc/rc*/*

  Any insights/hints for what I could try?

  As is, I can't reboot the machine remotely because it never gets to
  starting the network/sshd, which is highly annoying. I need to
  physically exit the maintenance shell.

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/util-linux/+bug/994749/+subscriptions