Hanno Foest wrote:
On Monday, Apr 07 2003 at 09:36:20 -0700, Samuel Flory wrote:
What you need to do is create an arrays for each drive in the promise bios. Each with only a single drive in it. This gives bios drives to boot off of.
Yes, it works indeed! Thanks! - But now my problem is... when I remove the hde drive, simulating a failure, the server doesn't boot from the remaining hdg drive, which at least partially thwarts the idea of increasing availability with RAID-1. (I'm using both drives in a Linux software RAID configuration which I boot of).
Any ideas?
Yes you need to install grub to both drives, or use lilo. The Red Hat installer doesn't install grub to all drives like it does for lilo. In the case of grub you need to do the following:
1)edit /boot/grub/device.map to look like this: (fd0) /dev/fd0 (hd1) /dev/hde (hd0) /dev/hdg
2)Install grub on /dev/hdg "grub-install /dev/hdg"
3)edit /boot/grub/device.map to look like this: (fd0) /dev/fd0 (hd0) /dev/hde (hd1) /dev/hdg
4)Install grub on /dev/hde "grub-install /dev/hde"
Lilo on the other hand should be configured correctly by the installer.
-- There is no such thing as obsolete hardware. Merely hardware that other people don't want. (The Second Rule of Hardware Acquisition) Sam Flory <sflory@xxxxxxxxxxxx>
