I was doing some testing earlier this week on RedHat 6.2 (2.2.14-5
recompiled w/o raid modules) using /boot RAID-1 (two disks, hda and hdc) and
saw what sounds like the same problem discussed here.
With hda unplugged, I would get 0x80 errors in lilo and was unable to boot
except by floppy.
I found that adding a "default = xxxxx" line to lilo.conf seemed to fix the
problem. Apparently letting lilo fall through to the first entry doesn't
properly write/generate the boot file on both disks.
-----Original Message-----
From: Chance Reschke [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, May 10, 2000 10:10 AM
To: [EMAIL PROTECTED]
Cc: Martin Munt; 'Richard Bollinger'
Subject: RE: Lilo RAID-1 woes
On Wed, 10 May 2000, Martin Munt wrote:
> Rich/Chance,
>
> I had EXACTLY the same problem. This post prompted me to fix it.
> Thanks a lot.
[ ... ]
I followed Richard Bollinger's advice, reordering the disks in my root
RAID-1 then rerunning lilo. With sda1 now raid-disk 0 and sdb1
raid-disk 1 the machine boots normally.
At Richard's suggestion I tested the integrity of the data lilo had
written to the second RAID disk by shutting down, removing sda1, and
rebooting. Consistent with the behavior we saw when the order of the
disks in my array was reversed, lilo complained with the boot-time
message "Error 0x80" and required me to enter a valid kernel label to
boot. This confirms that lilo is not properly updating the second
disk in the array.
- C
--
Chance Reschke
UW Astronomy
