LILO with root-raid nightmare - please help.
Hello,
I've replaced a failed disk in my raid 1 setup. I replaced the disk (hda
on-board ide), however the system then hung on reboot, this was because
hda, being blank, had no MBR. The second disk (hde) is plugged into a
promise ide card. So I went into the bios and set it to boot off the
promise card first. Reboot when fine. I partitioned the new disk and
added it to the array.
Then I tried to run lilo. Both disks need a valid mbr if I'm to be able
to boot in the event of a failed disk (the raid-extra-boot line in
lilo.conf). However this is what lilo said:
LILO version 22.6.1 (test mode), Copyright (C) 1992-1998 Werner Almesberger
Development beyond version 21 Copyright (C) 1999-2004 John Coffman
Released 17-Nov-2004, and compiled at 12:32:32 on May 25 2005
Debian GNU/Linux
Warning: Kernel & BIOS return differing head/sector geometries for
device 0x80
Kernel: 13872 cylinders, 16 heads, 63 sectors
BIOS: 1023 cylinders, 255 heads, 63 sectors
Using BIOS device code 0x80 for RAID boot blocks
Reading boot sector from /dev/md1
Fatal: Filesystem would be destroyed by LILO boot sector: /dev/md1
I believe the "Warning" line may be my bios switching the drive
addresses in order to boot off the promise card (is this correct?).
However I can't switch things back until I have a valid mbr on hda. I've
no idea what the "Fatal:" line means. md1 is the boot partition made up
of hda1 and hde1. My lilo.conf is as follows:
lba32
boot=/dev/md1
raid-extra-boot=/dev/hda,/dev/hde
install=/boot/boot-menu.b
map=/boot/map
delay=120
vga=normal
default=RAID
image=/boot/vmlinuz-2.4.24.raidserver1
label=RAID
read-only
root=/dev/md3
Is it safe to just put the disk with a valid mbr (hde) onto the on-board
controller? This would appear to be a simple way around the issue but
I'm not sure if this would break the array in some other way.
Any help solving this would be much appreciated.
regards,
drbob
Reply to: