[aklug] Apparent failure of mind - lilo won't boot

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ok, I've been doing this (linux/*BSD) a while, and am not exactly dumb
(although I do dumb *things* sometimes ;), so this is driving me nuts.
I realize this is probably something simple I'm missing, which is why
I'm here (if I thought this was a problem with LILO I'd be in that
mailing list ;).

Small background: Recently my debian testing installation on my home
backup server stopped booting. Short version is grub2 decided my
core.img was too big after an update from afar (deployed), shortly
thereafter we had a power outage and the system (of course) would not
boot. This system is not critical for day-to-day use, so it sat until I
returned.

Upon return, I decided to use this excuse to simplify the installation.
I wanted to use crux [1] as the OS, and *really* trim down the excess
fat in the install. As an aside, all my data is on `/dev/sdb` (one half
of the RAID-1 array which used to be active). I'm using `/dev/sda` for
this conversion, after which I'll rescue the data from `sdb`.

[1]: http://crux.nu

Unfortunately, I appear to have run into a wall I can't stop hitting
with my head. I've tried multiple different methods of install, and
finally decided I need help... so here I am.

Current results:

Unable to get lilo to boot with configuration described below, or other
similar configurations where `/` and/or `/boot` are on a RAIDed
partition (tried both).

Current configuration (all filesystems are xfs):

- - /boot (xfs) on RAID-1 device (/dev/md0)
- - swap on RAID-1 device (/dev/md1)
- - LVM VG named NAS on RAID-1 device (/dev/md2)
  - NAS partitioned into /, /srv/reverse, /srv/family

Linux kernel version is 2.6.35.6 (tried with current stable, no
difference). Lilo version is 23.2.

Error message is as follows.

#v+
LILO Loading CRUX............................
BIOS data check successful
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)
Pid: 1, comm: swapper Not tainted 2.6.35.6 #1
Call Trace:
 [<c11efd7a>] ? panic+0x4a/0xaf
 [<c12d4ad2>] ? mount_block_root+0x1d5/0x1e9
 [<c12d4be4>] ? mount_root+0x39/0x4d
 [<c12d4d3a>] ? prepare_namespace+0x142/0x168
 [<c12d485e>] ? kernel_init+0x195/0x1a0
 [<c12d46c9>] ? kernel_init+0x0/0x1a0
 [<c1002a56>] ? kernel_thread_helper+0x6/0x10
#v-

I've tried metadata version 0.9 and 1.0 for the md devices, with no
luck. I've ensured the kernel configuration has built in support for
xfs, RAID (md) devices, LVM, the SATA_SIL driver (for my Silicon Image
3114 SATA controller), and Intel IDE drivers for my IDE CDROM drives.
My lilo.conf is below, and I use lilo -H -F to make sure both /dev/md0
and /dev/sda are updated.

#v+
#
# /etc/lilo.conf: lilo(8) configuration, see lilo.conf(5)
#

lba32
install=text
boot=/dev/md0
raid-extra-boot=/dev/sda
image=/boot/vmlinuz
        label=CRUX
        root=/dev/NAS/root
        read-only
        append="quiet"

# End of file
#v-

At this point I'm at a stand-still, although I know a configuration like
this has worked in the past: This same system was an entirely LVM on top
of RAID solution until a few months ago. So does anyone have any
suggestions? Or just directions to look? Also, if anyone needs more
information to make a WAG, let me know and I'll provide.

Regards,
- --
dave [ please don't CC me ]
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (Darwin)

iEUEARECAAYFAk4abE0ACgkQzahokXOb2UzxmwCgkRdQNzJolVThyzKXSfz0V3DE
V5cAmLXKbyKKdlYyWDtiQoiRjp3hBSU=
=9lpd
-----END PGP SIGNATURE-----
---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.
Received on Sun Jul 10 19:22:01 2011