Ubuntu
linux-source-2.6.22 package

can't mount secondary sata drive

Bug #153027 reported by Jan Pfeifer on 2007-10-15

Affects		Status	Importance	Assigned to	Milestone
	evms (Ubuntu)	Invalid	Undecided	Unassigned
	linux-source-2.6.22 (Ubuntu)	Invalid	Undecided	Unassigned

Bug Description

Binary package hint: linux-image-2.6.22-14-generic

After upgrading from feisty to gutsy, the new kernel, 2.6.22-14 just won't mound my SATA "sdb" partitions. Mount complains that the disk is "busy".

Rebooting with the old kernel (2.6.20), that i luckily didn't remove, they mount fine.

My /etc/fstab:

# <file system> <mount point> <type> <options> <dump> <pass>
proc /proc proc defaults 0 0
/sys /sys sysfs defaults 0 0
# /dev/sda3
UUID=61d708a4-f6ea-4526-8917-22f8d613517c / ext3 defaults,errors=remount-ro 0 1
# /dev/sdb1
/dev/sdb1 /srv ext3 defaults,errors=remount-ro 0 0
# /dev/sda2
UUID=3d54a981-049d-45dc-987c-aadd398f4760 none swap sw 0 0
/dev/cdrom /media/cdrom0 udf,iso9660 user,noauto 0 0

# Windows disks
/dev/sda1 /media/win ntfs ro,defaults,umask=0222 0 0
/dev/sdb2 /media/fat vfat rw,defaults,umask=0,uid=65534,gid=65534 0 0
/dev/sdb3 /media/ntfs ntfs rw,defaults,umask=0222,uid=65534,gid=65534 0 0

Revision history for this message

Jan Pfeifer (pfjan) wrote on 2007-10-15:

Extract from /var/kern.log of the last time i tried to boot from kernel 2.6.22-14 Edit (49.5 KiB, text/plain)

Revision history for this message

Jan Pfeifer (pfjan) wrote on 2007-10-16:

Found the problem: EVMS was moving, or acquiring the /dev/sdb1 device, and creating another node under /dev/evms/sdb1.

I'm not sure if i had EVMS installed under Feisty, but either upgrading it to Gusty or having it installed when upgrading to Gusty triggered the problem.

Took me almost an hour (rebooting the machine) to find out the problem, and i have quite a good understanding of linux. Others might be left at a total loss. So I would strongly suggest that the EVMS package first check /etc/fstab for the partitions it takes over, before it moves it for somewhere else. And either avoid taking over them in that case, or printing warning messages with the problem.

Is there a way to move this bug from linux-source-2.6.22 to evms ?

Thanks a lot!!! Sorry for the wrong classification of the bug -- i didn't know then.

Changed in linux-source-2.6.22:
status:	New → Invalid

Revision history for this message

Oisín Mac Fhearaí (denpashogai) wrote on 2007-10-20:

Thanks for this - I didn't know EVMS made them available elsewhere.
But Ubuntu seems to automatically use disk UUIDs rather than /dev/sdaX-style partition references. And after upgrading to Gutsy, my UUIDs still point to the old devices.

Shouldn't they be updated when the evms package is installed..?

Revision history for this message

Jan Pfeifer (pfjan) wrote on 2007-10-21:

You are right about Ubuntu using the UUIDs to mount by default. What happened in my case is that the drive was installed later (after Ubuntu was already installed)... so i simply formatted it (ext3) and mounted it myself, using the /dev/sdXn style partition reference.

i was thinking, if /etc/fstab is changed to the /dev/evms/sdXn format, it will no longer work in the previous version of the kernel (feisty) ... so one of the systems (feisty or gutsy) will be broken anyway ... it seems to me that ideally the package should alarm at boot time when it sees an entry in /etc/fstab with an invalid device, that EVMS took over. And maybe have fstab changed automatically to the disk UUIDs fstab format ...

ideas?
thx :)

Revision history for this message

Jeff Anderson (jander99) wrote on 2009-07-20:

Hello Jan and thank you for filing a bug report. Unfortunately this issue has not seen attention in quite some time. Currently evms is only available in Dapper and Hardy. If you're still experiencing this issue please let us know however be aware evms has been dropped from Debian and most likely will not be used in Ubuntu. Thanks!

Changed in evms (Ubuntu):
status:	New → Incomplete

Revision history for this message

xteejx (xteejx) wrote on 2010-10-13:

We are closing this bug report because it lacks the information we need to investigate the problem, as described in the previous comments. Please reopen it if you can give us the missing information, and don't hesitate to submit bug reports in the future. To reopen the bug report you can click on the current status, under the Status column, and change the Status back to "New". Thanks again!