hardy upgrade: /dev/disk/by-uuid non existant at boot; dropping to a shell (with "386" kernel)

Bug #213884 reported by Pietro Battiston
16
This bug affects 1 person
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

Binary package hint: linux-image-2.6.24-15-386

I just upgraded a desktop PC from Gutsy to Hardy (without errors in upgrading).

I found two new kernels in Grub's list: with the second one (2.6.24-15-generic), everithing goes smooth.

Instead, with the first one, 2.6.24-15-386, system doesn't boot: in normal mode I just see the splash orange bar going left and right, while in recovery mode it hangs on "Begin: Waiting for root file system"... but after some time (in both modes) I get back to text mode and see:

"Check root= bootarg cat /proc/cmdline
or missing modules, devices: cat /proc/modules ls /dev
ALERT! /dev/disk/by-uuid/3d866059-4b4c-4c71-a69c-213f0e4fbf32 does not exist. Dropping to a shell!"

and then "BusyBox..."

In the shell I get, I'm not able to save dmesg output (but I cheched that nothing appears if grepping with "sd", "scsi", and "hd"), while uname -a says:
Linux (none) 2.6.24-15-386 #1 Fri Apr 4 03:10:49 UTC 2008 i686 Unknown

I also checked if the "/dev/disk..." think existed (I checked it does exist when I boot with "generic"), but the directory "/dev/disk" doesn't exist at all (neither does any "/dev/sd*" or "/dev/hd*").

Here are the corresponding /etc/fstab lines:
# /dev/hda4
UUID=3d866059-4b4c-4c71-a69c-213f0e4fbf32 / ext3 defaults,errors=remount-ro 0 1

Some apparently non-important (but who knows) things:
- this is a manually edited /etc/fstab, but this line was automatically created (replaced), I don't know if in base (Feisty) install or in some upgrade.
- in upgrading to Hardy all my "/dev/hd*" devices become "/dev/sd*"
- this is a system installed as Ubuntu but now used as Xubuntu (by installing the related metapackage).

I'm attaching "sudo lscpi -vvnn" output.

Tags: cft-2.6.27
Revision history for this message
Pietro Battiston (toobaz) wrote :
Revision history for this message
Pietro Battiston (toobaz) wrote :

This is the dmesg output WITH THE WORKING ("generic") KERNEL.

Revision history for this message
Tony Green (ubuntu-beermad-deactivatedaccount) wrote :

I have had a similar experience with upgrading to Hardy.

The new kernel managed to mount my root filesystem (/dev/hda1) but failed to mount any more of them. running "fdisk /dev/hda" and "fdisk /dev/hdb" I received an error message that the devices did not exist.

In addition, either the upgrade or the new kernel corrupted the superblock of EVERY filesystem on both of my hard discs - booting the first time with the new kernel, it repaired the root filesystem before failing as above. When I rebooted using the old kernel, it had to repair the superblock on every other filesystem before coming up.

Of course, because of the old kernel expecting a different Nvidia driver, X was unusable with this kernel, so I have restored my system from the pre-upgrade backups. I have however kept a dump of / /usr and /var from the failed install, so I can get any information from this that may be needed to resolve this issue.

One note about my upgrade. I was unable to use the method recommended at https://help.ubuntu.com/community/HardyUpgrades/Kubuntu as Adept didn't get the expected upgrade button (see https://bugs.launchpad.net/ubuntu/+source/adept/+bug/219637) so upgraded as recommended by a respondent to that bug by running "sudo do-release-upgrade -d" which did the whole upgrade for me.

Revision history for this message
Tony Green (ubuntu-beermad-deactivatedaccount) wrote :

Just noticed the title of this bug specified "386" kernel. My problem was with an AMD 64 kernel.

Revision history for this message
Dave Vree (hdave) wrote :

I can confirm this issues. In my particular case I found out what the problem was and fixed it. Perhaps my problem is only one way to reproduce this bug, but in any event...here's the issue/fix:

The hard drive UUID in my GRUB menu.lst was wrong. I had swapped out my hard drive when my machine was running Gutsy. I had fixed the UUID in /etc/fstab (you have to or you can't login). However I never changed the UUID in /boot/grub/menu.lst and when I upgraded to Hardy I used the new "merge" capability to get my new menu.lst file.

Apparently Gutsy didn't care the UUID was wrong, but Hardy does. I booted into windows, copied the UUID from fstab and pasted it into menu.lst I now I am swimming in 8.04 goodness.

Revision history for this message
Tony Green (ubuntu-beermad-deactivatedaccount) wrote :

In my case, the UUID in menu.lst matches what the entry in /dev/disk/by-uuid links to my root filesystem.

BUT... /etc/fstab refers to filesystems by the much more sensible (and memorable) /dev/hd* names; I think because I edited it back that way after an earlier upgrade. I'll have to try putting UUIDs back in and see if it makes a difference.

I'm sure there must be a reason for Ubuntu's insistence on using UUIDs rather than /dev/hd* names, but I'm sure it's outweighed by the inconvenience - after all, what's easier to remember, /dev/hda1 or c9593a22-7236-4321-9fe3-a5187ed87c53? I know which I find easier.

Revision history for this message
Tony Green (ubuntu-beermad-deactivatedaccount) wrote :

I can confirm that the reason for my problem was, as I suggested, that I was using /dev/hd* references in /etc/fstab.

I have changed it to use UUIDs and installed the now-released Hardy Heron and the system boots OK now.

Revision history for this message
L_Mo (lance-mosher) wrote :

I can confirm this for kernel 2.6.24-16.

I attempted to boot the Hardy liveCD, but it would not load the kernel. The error was: '/sbin/modprobe' abnormal exit, followed by the BusyBox shell.

I successfully installed linux using Gutsy, then upgraded using do-release-upgrade. I can boot this with no problem using the 2.6.22 kernel, but the 2.6.24 kernel fails as described by the original poster.

I double-checked. My /etc/fstab is in UUIDs, and grub's menu.lst matches.

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

The Ubuntu Kernel Team is planning to move to the 2.6.27 kernel for the upcoming Intrepid Ibex 8.10 release. As a result, the kernel team would appreciate it if you could please test this newer 2.6.27 Ubuntu kernel. There are one of two ways you should be able to test:

1) If you are comfortable installing packages on your own, the linux-image-2.6.27-* package is currently available for you to install and test.

--or--

2) The upcoming Alpha5 for Intrepid Ibex 8.10 will contain this newer 2.6.27 Ubuntu kernel. Alpha5 is set to be released Thursday Sept 4. Please watch http://www.ubuntu.com/testing for Alpha5 to be announced. You should then be able to test via a LiveCD.

Please let us know immediately if this newer 2.6.27 kernel resolves the bug reported here or if the issue remains. More importantly, please open a new bug report for each new bug/regression introduced by the 2.6.27 kernel and tag the bug report with 'linux-2.6.27'. Also, please specifically note if the issue does or does not appear in the 2.6.26 kernel. Thanks again, we really appreicate your help and feedback.

Revision history for this message
davidrachlin (davidrachlin) wrote :

Leann...

I am interested in testing the new Kernel for the purposes of using my GPS. I have a Delorme LT-40 which is supposed to work with the newest kernel, but which does not work at all with my 2.6.24-19-generic x86 AMD64 kernel.

If I download the live CD and install it on a partition on my drive... how can I later go back and delete that partition without losing all my extant data?

Thanks

(I'm still a relatively new user, and not super familiar with Linux yet, but I want to learn desperately!!!)

Revision history for this message
Charlie Kravetz (cjkgeek) wrote :

This bug report is being closed due to your last comment regarding this being fixed with a configuration change. For future reference you can manage the status of your own bugs by clicking on the current status in the yellow line and then choosing a new status in the revealed drop down box. You can learn more about bug statuses at https://wiki.ubuntu.com/Bugs/Status . Thank you again for taking the time to report this bug and helping to make Ubuntu better. Feel free to submit any future bugs you may find.

davidrachlin - please use the mailing list (info at https://lists.ubuntu.com/mailman/listinfo/ubuntu-users) or #ubuntu-testing to ask your questions. Thanks.

Changed in linux:
status: New → Invalid
Revision history for this message
Fixxser (steve-333) wrote :

For me, this was the problem I had. Just installed Ubuntu 8.10 - the Intrepid Ibex on a Intel base.
There was a big delay on booting from bios to os as well. Turned out I had a jumper on the extra data drives I had in the system, (Western Digital), on the master w/slave jumpers. Diagram showed for single or master it should be all NO jumpers, blank. It works and boots great now.

Revision history for this message
gyurman (j-launchpad-20-gyurman-spamgourmet-com) wrote :

Only need reinstall grub2 and uninstall grub and these kernels which aren't refresh. This is good for me.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.