RE: disk arms (was RE: Tips for user ASP)

>From: "Dan Bale" <dbale@samsa.com>
>
>With all due respect, I find it hard to believe that a 4gb drive at 90%
>capacity could perform significantly better than an 8gb drive at 45%
>capacity, raided or not.  How could that happen?
>
> > But I don't really think very many folks are using RAID, are they?
>
>I think you were TIC here, right?  AFAIK, I’m sure most shops are using
>RAID, and if you’re not counting “mirrored” RAID, you’re still talking
>significant numbers.  At least in my experience.
>

I _might_ have been TIC.  Just maybe.

I am short on time.  More later...

Currently, RAID information is put on one "end" of the disk.  Inner ring or
outer ring, I don't know.  "Real" data (mine and yours) gets put on the
other.  So the only thing that is between RAID bits and real bits is the
empty space.  The empty space is bigger on empty disks.

By far, the slowest part of any disk operations are the "seek times" which
move the head from one cylinder to another, inner ring to outer ring or
reverse.  A RAID implementation sort of guarantees that you will be moving
the arm between RAID bits and real bits, wherever they are on the disk.  So
empty RAID disks have a lot longer seek times than almost full RAID disks.
Changing the rotational speed from 5400 rpm to 10,000 rpm does not change
the seek time, it only improves the rotational delay time, a different
component of disk performance.  I have heard that IBM may move the location
of the RAID bits to improve this unfortunate situation, but I don't believe
it has been announced.

--  Charly



"Nothing would please me more than being able to hire ten programmers and
deluge the hobby market with good software."  -  Bill Gates in 1976

"We are still waiting..."  -  Alan Cox in 2002

"Linux is only free if your time is worthless."


Charly Jones
253 265-6244
Gig Harbor
Washington USA



_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com