/dev/md0 construct not supported in Feisty

Bug #85822 reported by Yakov Dekel
28
Affects Status Importance Assigned to Milestone
debian-installer (Ubuntu)
Fix Released
Undecided
Unassigned
partman-md (Ubuntu)
Fix Released
Undecided
Unassigned

Bug Description

During install, a raid was not recognized. Apparently the installer searches for /dev/m/0 where the existing raid is defined to reside on /dev/md0.

"ln /dev/md0 /dev/md/0 -s" enabled me to continue with the install successfully.

Further on, the system wont boot. At thie stage I generated an mdadm.conf which seemed to force generating the raid.

In Dapper and Edgy these operations and especially the creation of mdadm.conf where not necessary. From looking at the log it seems that /dev/md0 WAS created during the kernel boot but was closed in favor of a "/dev/md/n" construct (which might be good for newly generated raids but not for existing ones).

ProblemType: Bug
Date: Sat Feb 17 18:52:37 2007
DistroRelease: Ubuntu 7.04
Uname: Linux study-linux 2.6.20-8-generic #2 SMP Tue Feb 13 05:18:42 UTC 2007 i686 GNU/Linux

Yakov Dekel (y1-dekel)
description: updated
Revision history for this message
Brian Murray (brian-murray) wrote :

Thanks for your bug report. Did this occur with the Herd 4 release of Feisty? Thanks in advance.

Revision history for this message
Yakov Dekel (y1-dekel) wrote : Re: [Bug 85822] Re: /dev/md0 construct not supported in Feisty

Brian Murray wrote:
> Thanks for your bug report. Did this occur with the Herd 4 release of
> Feisty? Thanks in advance.
>
> ** Changed in: Ubuntu
> Assignee: (unassigned) => Brian Murray
> Status: Unconfirmed => Needs Info
>
That is correct. Herd 4 - alternate.

Regards,
Yakov.

Revision history for this message
arcterex (alan-ufies) wrote :

As a note this may still is happening in herd5 - alternate. An install on root raid after initial reboot drops down to initramfs with the 'no job control' error message. I fixed this by following the instructions at http://ubuntuforums.org/showthread.php?p=2272436

Not sure if this is the exact same issue. However, it appears from other bugs here that feisty and root raid aren't mixing well :(

Revision history for this message
Yakov Dekel (y1-dekel) wrote :

Hi,
Yes, Herd5 is also broken, and even the workaround I had for Herd4 does
not work this time, because the mere existence of a raid partition on
the disks causes Herd5 installer to say "boot partition not found" (and
I have more than a dozen partitions on my two drives, only two of them
are raid).

Thanks for the update.

arcterex wrote:
> As a note this may still is happening in herd5 - alternate. An install
> on root raid after initial reboot drops down to initramfs with the 'no
> job control' error message. I fixed this by following the instructions
> at http://ubuntuforums.org/showthread.php?p=2272436
>
> Not sure if this is the exact same issue. However, it appears from
> other bugs here that feisty and root raid aren't mixing well :(
>

Revision history for this message
Yakov Dekel (y1-dekel) wrote : Re: /dev/md0 construct not supported in Feisty - Beta

Update from beta installation attempt - failure.

The raid, which is configured as /dev/md0 is ignored, and the installer is consistently looking for /dev/md/* which it cannot find. "ln /dev/md0 /dev/md/0 -s" did the trick.

BUT

lvm is now broken. My lvm sits on top of the raid. I can define a new logical volume, but it can not be selected from the list of partitions. ALL logical volumes are listing correctly, but selecting any one of them (including, of course the newly created one) returns to the same screen. There is NO message in the log screen when this happens (alt-f4).

Yakov.

Revision history for this message
Giovanni Premuda (gpremuda) wrote :

Even without LVM "ln /dev/md0 /dev/md/0" satisfies the installer, so that i happily arrives to completion, but the system doesn't boot, it throws you to a (frozen? or is it my USB keyboard?) busybox shell with error "/bin/sh:can't access tty; job control turned off"

Revision history for this message
Samuel Hangouët (samuel-hangouet) wrote :

Well, just to tell I encountered the same problem during ubuntu installation from the last hurd alternate CD.

The system could not boot on a soft RAID0 array of two partitions located on SATA disks,
whereas installation without RAID succeeded.

I also noticed the lack of /dev/md0 from the busybox shell (inside initrd I guess).

..And a lot of ubuntu users seem to have the same problem, as noticed in this thread :
http://forum.ubuntu-fr.org/viewtopic.php?pid=861296

Revision history for this message
Andreas Moog (ampelbein) wrote :

Thank you for taking the time to report this bug and helping to make Ubuntu better. You reported this bug a while ago and there hasn't been any activity in it recently. We were wondering is this still an issue for you? Can you try with latest Ubuntu release? Thanks in advance.

Changed in partman-md:
assignee: nobody → andreas-moog
status: New → Incomplete
Changed in debian-installer:
status: New → Incomplete
Revision history for this message
Yakov Dekel (y1-dekel) wrote :

I can confirm that this does not happen in Hardy. I guess this can be closed.

Revision history for this message
Andreas Moog (ampelbein) wrote :

This bug report is being closed due to your last comment regarding this being fixed with an update. 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.

Changed in partman-md:
assignee: andreas-moog → nobody
status: Incomplete → Fix Released
Changed in debian-installer:
status: Incomplete → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

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