Intel ICH8R chipset

[Hall, Eric R]

I agree.  BIOS hasn't a clue about the raid configuration.  But in
regards to md, couldn't an initrd be used to mount "/" on the raid from
the next disk from that raid?

Once md is loaded at initrd, you would then have access to the raid and
would be able to continue boot.  It'll be in degraded mode, but it
should boot.  (I could be very wrong though!)



-----Original Message-----
From: Daniel O'Connor [mailto:darius at dons.net.au] 
Sent: Wednesday, February 28, 2007 9:46 PM
To: Hall, Eric R
Cc: ATARAID (eg, Promise Fasttrak, Highpoint 370) related discussions;
Darrick J. Wong
Subject: Re: Intel ICH8R chipset

On Thursday 01 March 2007 15:39, Hall, Eric R wrote:
> Or one could just mirror /boot from the first disk to the first two or
> three disks.  It'd be a pain, but you might still be able to boot...

How? The BIOS is doing the reading from those disks if it doesn't
understand 
how to use the redundancy there is a problem.

I would be surprised if a "firmware" RAID system (like say a Promise
FT100TX2 
or FT2300) would be any slower than using those disks on a comparable
but 
non-RAID controller.

Once the OS starts it is its responsibility to maintain the metadata but
it 
should still be able to get a good speed. I would say the requirement to
sync 
the metadata to disk will slow it down but a pure software RAID system
should 
do the same or risk array corruption.

-- 
Daniel O'Connor software and network engineer
for Genesis Software - http://www.gsoft.com.au
"The nice thing about standards is that there
are so many of them to choose from."
  -- Andrew Tanenbaum
GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C