No error message displayed when device fails to mount on Ubuntu Server 10.04

[CharlesA]

This has only happened in Lucid as far as I know, it didn't happen in Karmic.

If a device listed in fstab is not found, the machine will hang until you press "s" to skip that device.

No error message is displayed, which is the problem.

I've had this happen on kernel version 2.6.32-21 through 2.6.32-24.

Basically what happens is when you add a drive/device to fstab and then the device fails or is removed, the OS just sits there waiting for user input without displaying an error message.

The only thing it says is:
fsck from util-linux-ng 2.17.2
/dev/sda1: clean, xx/xxxxx files, xx/xxxxxx blocks

After you hit "s" to skip the device, it proceeds to boot normally after displaying "skipping /mountpoint at user request

ProblemType: Bug
DistroRelease: Ubuntu 10.04
Package: linux-image-2.6.32-24-server 2.6.32-24.38
Regression: No
Reproducible: Yes
ProcVersionSignature: Ubuntu 2.6.32-24.38-server 2.6.32.15+drm33.5
Uname: Linux 2.6.32-24-server x86_64
AlsaDevices: Error: command ['ls', '-l', '/dev/snd/'] failed with exit code 2: ls: cannot access /dev/snd/: No such file or directory
AplayDevices: Error: [Errno 2] No such file or directory
Architecture: amd64
ArecordDevices: Error: [Errno 2] No such file or directory
CurrentDmesg: [ 15.191519] eth0: no IPv6 routers present
Date: Sun Jul 25 14:23:45 2010
InstallationMedia: Ubuntu-Server 10.04 LTS "Lucid Lynx" - Release amd64 (20100427)
Lsusb:
 Bus 002 Device 002: ID 80ee:0021
 Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
MachineType: innotek GmbH VirtualBox
PciMultimedia:

ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.32-24-server root=UUID=97e563d6-4a5b-4bda-8686-a44b00161a96 ro quiet
ProcEnviron:
 LANG=en_US.UTF-8
 SHELL=/bin/bash
SourcePackage: linux
dmi.bios.date: 12/01/2006
dmi.bios.vendor: innotek GmbH
dmi.bios.version: VirtualBox
dmi.modalias: dmi:bvninnotekGmbH:bvrVirtualBox:bd12/01/2006:svninnotekGmbH:pnVirtualBox:pvr1.2:
dmi.product.name: VirtualBox
dmi.product.version: 1.2
dmi.sys.vendor: innotek GmbH

[CharlesA (charlesa)]

Having no error message like the one displayed in Ubuntu Desktop causes confusion as to what is wrong.

See attached screenshots for what happens if the OS is unable to mount a device listed in fstab on boot.

[CharlesA (charlesa)]

After hitting "s" the OS boots normally.

[Jeremy Foshee (jeremyfoshee)]

Hi Charles,

If you could also please test the latest upstream kernel available that would be great. It will allow additional upstream developers to examine the issue. Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Once you've tested the upstream kernel, please remove the 'needs-upstream-testing' tag. This can be done by clicking on the yellow pencil icon next to the tag located at the bottom of the bug description and deleting the 'needs-upstream-testing' text. Please let us know your results.

Thanks in advance.

    [This is an automated message. Apologies if it has reached you inappropriately; please just reply to this message indicating so.]

[CharlesA (charlesa)]

Installed /~kernel-ppa/mainline/v2.6.33.5-lucid
2.6.33-02063305-generic x86_64

Same thing - no error message displayed.

Installed /~kernel-ppa/mainline/v2.6.34-lucid
2.6.34-020634-generic x86_64

Same thing - no error message displayed.

Installed /~kernel-ppa/mainline/v2.6.35-rc1-lucid
2.6.35-020635rc1-generic x86_64

Same thing - no error message displayed.

I did notice that the problem didn't appear to happen in Ubuntu Server 9.10 x64, with any kernel, it appeared to just skip the problem device.

Andreas mentioned something about Plymouth here: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/576001/comments/34

Not sure how that affects not being able to display an error message without having the splash screen.

[Gerald Villemure (gvillemure)]

The reason you don't see anything is that you disabled the Ubuntu splash screen.

If you remove the kernel options "quite splash" and you have entries in your fstab that the system has a problem mounting.
 **Your in trouble!**

I use a serial console to manage all my servers and this upstart bug is a major concern for me.

I may have to forgot upgrading my servers to Lucid because of this.

Gérald

[CharlesA (charlesa)]

Does the default install of Server even have the splash screen installed?

I am running a plain install of Lucid Server x64. I know the desktop version has the splash screen - plymouth, isn't it? I don't recall if it's installed and enabled by default on the server install.

I checked what the kernel boot options look like on a VM copy of my physical server and they are attached.

Here's what the grub options look like:

recordfail
insmod ext23
set root='(hd0,1)'
search --no-floppy --fs-uuid --set ea9998bb-6052-4a29-b588-525a7693f1bf
linux /boot/vmluinuz-2.6.32-24-server root=UUID=ea9998bb-6052-4a29-b588-525a7693f1bf ro quiet
initrd /boot/initrd.img-2.6.32-24-server

I removed the "quiet" part and it still got stuck after running fsck and sat there without displaying an error message.

Is there a way to force it to auto-skip devices that are not ready?

[CharlesA (charlesa)]

I see what you meant now, you need to "enable" the splash screen by adding "splash" after "quiet" in the boot options.

With that enabled, it displayed the error message. It would be nice to not have to rely on a splash screen to display error messages when it is not turned on by default.

I would much rather have it display the error on the terminal instead.

[Gerald Villemure (gvillemure)]

> I would much rather have it display the error on the terminal instead.

Tell me about it. I don't have monitors attached to the servers I admin so this is a real prob for me.

I can no longer admin the servers strictly from the serial console.

Gérald

[Gerald Villemure (gvillemure)]

You know what I would REALY love to see since I am always concerned about mountall giving me grief is and new option for the sixth field in fstab.

If would be great if mountall could be updated such that the sixth field in fstab is:
0 Don't check and **DON'T PAUSE**, filesystem does not need to be checked, if mount fails keep booting.
1 ROOT Filesystem
2 Other Filesystems, checked sequentially
3 Filesystem is checked/mounted in the background ideally in a "screen" session, most importantly it will NOT PAUSE the boot and the admin can check the progress with SSH and "screen".

Another route is to NOT mention any data Filesystems in the fstab at all and have them mounted with a custom script.

Gérald

[moli (f-launchpad-moli-hu)]

oh god. i had this bug right now, i had to get my server from colocation as sshd didnt start because the boot stopped at mount/fsck (i stupidly added an automount for a -missing- pendrive with fsck option then made system to force fsck at boot. Not that hard mistake to make as it sounds at first.)

please make mount and/or fsck AUTOMATICALLY SKIP missing devices.

[Gerald Villemure (gvillemure)]

Even if you set your entry in /etc/fstab to NOT run fsck, the current mountall will HANG your system if there is a problem.

BTW, I wrote a small script to make booting a system a little less scary. You can find it here, https://bugs.launchpad.net/bugs/614662

Gérald