[Leica] RAID controllers?

[hartzell at alerce.com (George Hartzell)]

Gary Todoroff writes:
 > Very thoughtful and informative reply, George. Would love to debate 
 > over a beer sometime! [...]

I think that it wouldn't be much of a debate, at the end of the day
we're pretty much in agreement.  Over the years we've all learned all
kinds of lessons about organizing our data on computers.  A lot of
what I've internalized isn't really appropriate anymore (multi core
systems with fast internal buses make it more reasonable to do
software raid, terabyte drives mean that I no longer have to glue
storage together out of too-small peices, although fast & large file
size cameras have me generating more data than ever) but there are new
reasons for slicing and dicing.

I was reacting to the idea of not putting the OS on a raid for
stability reasons, and I still think that's not an issue.  Your
comment 

 > [...] data only on 
 > the RAID, OS on the IDE (or is it SATA?) drive that runs the computer 
 > and will be there no matter what happens to the RAID.

still makes it seem that you don't see the RAID devices as a source of
stability, but we can just agree to differ on that.

 > I also like the 
 > performance advantage that puts the OS overhead on one disk, while 
 > the data read/writes are the only function of the RAID.

One of the nice things about putting everything on the mirror is that
it gives the OS a chance to interleave and nearly-optimally schedule
IO requests onto both of the drives, you get 2x read speeds while
reading in LightRoom for the first time, then 2x read speeds while
reading in your images, then if necessary app and image reading in
parallel.  

But anyway, it's been a fun discssion,

g.