Subject: Re: Removing "rows" from the RAIDframe driver..
To: Jason Thorpe <thorpej@wasabisystems.com>
From: Greg Oster <oster@cs.usask.ca>
List: tech-kern
Date: 12/20/2003 12:17:56
Jason Thorpe writes:
> On Dec 20, 2003, at 8:51 AM, Greg Oster wrote:
> 
> >  1) Anyone wishing to make multiple-row support fully functional will
> > have a LOT of work on their hands to re-add the row bits.
> 
> Does *anyone's* RAID implementation actually support multiple rows per 
> stripe?  Conceptually, it's even tricky to wrap your head around simply 
> because a single-row stripe is trivial to visualize or draw a diagram 
> of.

The "row" code in RAIDframe would behave, in theory, much like doing 
a concatenation ccd of RAID sets.  i.e. if there are n stripes in the 
first row, then stripe n+1 would be the first stripe of the 2nd row.
I forgot to mention, but had the row functionality actually worked in 
RAIDframe, we could *still* replace it with a concatenation ccd and 
regular "1D" (one-row) RAID sets.

I'm not sure if Veritas can do 2D arrays or not, but I think that was 
the example brought up the last time I floated the idea of nuking 
the row code.

> In other words, I'm not sure you're creating an actual problem by 
> removing that code :-)

I don't think I am either, but I figure it won't hurt much to get 
some second opinions. :)  

Later...

Greg Oster