Re: testing raid 0

[Jeff Bachtel <jeff@cepheid.org>]

Added layer of driver complexity, untuned raidframe config, and the
fact that the interleave on most drives is tuned for continuous
reads/writes across the surface, and not periodically interleaved r/w.

Those would be my guesses, but I am not a disk guru. And thats for
Raid 0. For Raid 1 you have an additional write operation. For Raid 5
you have additional parity calculations.

jeff

On Wed, Aug 01, 2001 at 08:55:17AM -0700, Chris Palmer wrote:
> On 01 Aug  Jeff Bachtel wrote:
> 
> > From personal experience with raidframe on OpenBSD, I'll second
> > Peter's statement. While it is theoretically possible that a
> > well-tuned raidframe setup (check the layout section of the raidctl
> > configuration, as well as the queue) could slightly outperform normal
> > disk drivers on a scsi chain, its not likely, and is even less likely
> > on an IDE chain.
> 
> Why is RAIDframe slower than a single disk?
> 
> 
> -- 
> Chris
> 

-- 
Jeff Bachtel  (root@ISC,TAMU)    http://www.cepheid.org/~jeff
				 [finger jeff@cepheid.org for PGP key]