Re: Oracle Benchmark Results for Different Hardware Configurations?

From: DA Morgan <damorgan_at_psoug.org>
Date: Fri, 15 Sep 2006 11:22:54 -0700
Message-ID: <1158344573.180633@bubbleator.drizzle.com>

HansF wrote:
> On Fri, 15 Sep 2006 07:00:53 -0700, DA Morgan wrote:
>

>> In the case of these trumped up benchmarks ... games are played by the
>> vendors that do not reflect anything remotely close to the real world.
>> If you tell me RDBMS alpha running on hardware platform beta with
>> operating system gamma gave a specific number that means nothing if I

>
> Although there is a certain amount of truth to what you say, I believe
> the full disclosures are an attempt to alleviate that issue.
>
> No matter what the scenario ... give those with a political interest an
> opportunity to manage or obfuscate the results to their advantage, and
> they will take that opportunity. For example, I've seen 'in house, based
> on the application' benchmarks become totally useless.
>
> One in particular comes to mind: management at a 'customer' site had an
> inhouse 'vendor x' specialist tune the application for 'vendor x' and for
> Oracle. The resulting benchmark and recommendations were ... interesting.

I am not surprised. I think any organization large enough to care about the hardware or RDBMS they are buying had best act in its own vested interest and do their own validation.

Internal developers may not know what they are doing on one platform rather than another and that will be reflected in the quality of their design and code. A third-party vendor such as SAP, Banner, PeopleSoft, Siebel ... let them prove themselves with different back-ends if they want the sale.

 From my experience Oracle has rarely been hurt by reality based testing.

-- 
Daniel Morgan
University of Washington
Puget Sound Oracle Users Group

Received on Fri Sep 15 2006 - 13:22:54 CDT