Re: asm on san
Date: Mon, 15 Dec 2008 12:42:20 -0800
Michael Austin wrote:
> The number of systems and applications that "require" the differences in
> those performance numbers vs. cost on a truly busy system are few and
> FAR between.
Sorry but that is truly nonsense on multiple levels.
Feel free to disagree if you wish but until you are prepared to present your findings at UKOUG or Collaborate with full documentation what you are offering is a personal opinion of which there is no scarcity.
> Based on experience, I would venture a guess, that the numbers produced
> from your lab environment would vary greatly not only dependent on the
> workload presented, but the hardware, hardware vendor and operating
> system used to generate those numbers. Not all systems are created
> equal. (Something most "commodity" vendors never get).
Of course they vary based on hardware and workload ... everything does. But RAID5 running well is essentially the same as saying I overbought hardware to make up for a lousy configuration. Anyone with deep enough pockets can do that but few CIOs will pay you to do so with their budget.
> All I am saying is that unless you actually DESIGN performance into the
> system**, your RAID setting is of very little consequence in 99.9% of
> the databases in use today
Again, sorry, but this is total nonsense. You can not increase writes by a factor of 2+ and have no consequences.
If you have the numbers write a White Paper. If not please do not peddle personal opinion as fact.
-- Daniel A. Morgan Oracle Ace Director & Instructor University of Washington damorgan_at_x.washington.edu (replace x with u to respond) Puget Sound Oracle Users Group www.psoug.orgReceived on Mon Dec 15 2008 - 14:42:20 CST