[Wolves] Chicken-and-egg situation regarding HPT372 RAID controller

Daniel Jones wolves at mailman.lug.org.uk
Thu Oct 10 01:22:00 2002

On Monday 07 October 2002 6:57 pm, you wrote:
> > Trouble is, the one drive on a conventional disk controller failing
> > would make the system unbootable in that instance.  Bad.
> You could boot the system from a ram disk, insert the required module,
> then pivot_root onto a new root filesystem on your RAID array.  I use
> something like this technique to run my root filesystem under LVM.  Since
> you're getting your kernel loaded as it is, the boot loader must recognise
> the RAIDed volumes, and you'll be able to load the ram disk from /boot
> without problems.
> Look at initrd(4) and linux/Documentation/initrd.txt for more info.

Peter, you, sir, are a total star.

The RAID chip is now *this close* to working right.

The only problem now is, though, that the thing sticks at the partition 
check.  Because it can still see hde-h, it tries (and fails) to check the 
partitions on hde.  Grrrrrrrrrrr.