[Virtual-pkg-base-maintainers] Bug#781172: base: Stock Debian Jessie - can't boot off degraded raid1?

carl c_boben at hotmail.com
Wed Mar 25 16:32:46 UTC 2015


Package: base
Severity: important

Dear Maintainer,


Install a Wheezy VM using debian-7.8.0-amd64-netinst.iso 
create a raid1 mirror in the drive setup
mount / on /dev/md0 and install base system to it
login
grub-install /dev/vdb
poweroff
disable either drive /dev/vda or /dev/vdb (give VM access to only 1 drive)
computer boots

# cat /proc/mdstat
Personalities : [raid1]
md0 : active raid1 vda1[1]
 8382400 blocks super 1.2 [2/1] [_U]

Sweet.



Install a Jessie VM using debian-jessie-DI-rc1-amd64-netinst.iso
create a raid1 mirror in the drive setup
mount / on /dev/md0 and install base system to it
login
grub-install /dev/vdb
poweroff
disable either drive /dev/vda or /dev/vdb (give VM access to only 1 drive)

ALERT! /dev/disk/by-uuid/XXXX does not exist. 
Dropping to a shell!

(where XXXX is the uuid of /dev/md0)

# cat /proc/mdstat
Personalities :
md0 : inactive vda1[0](S)
 8382464 blocks super 1.2

Both refer to md0 by UUID in fstab.

When I remove the "quiet" from the kernel boot args, this is the last line I see repeated over and over again until finally I get the error I described above:

running /scripts/local-block
 
I'm admitedly not a raid expert, but this seems to be really missing the point. Part of the reason I want a raid1 is so it can survive this type of scenario. 

I googled around and surprisingly came up with very few leads. I found a Ubuntu wiki (https://help.ubuntu.com/community/Installation/SoftwareRAID#Boot_from_Degraded_Disk) that discussed how to force it to boot a degraded raid but seemed to do nothing here on debian. I found a kernel boot arg that promised to do the same (http://serverfault.com/questions/196445/boot-debian-while-raid-array-is-degraded), but it didn't work either.

What am I missing here? Shouldn't this work by default? I think a lot of people will get caught off guard by this thinking their startup drive is mirrored but the server won't boot without some manual intervention? Actually, I was not able to get it to boot at all.




-- System Information:
Debian Release: 7.8
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: amd64 (x86_64)

Kernel: Linux 3.16.0-0.bpo.4-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_CA.UTF-8, LC_CTYPE=en_CA.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash



More information about the Virtual-pkg-base-maintainers mailing list