Subject: | |
From: | |
Reply To: | |
Date: | Thu, 8 Mar 2001 15:22:18 EST |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Jim writes:
> On 8 Mar 2001, at 10:57, Mark Boyd wrote:
>
> > All questions regarding performance issues are met with nebulous
> > and meaningless responses or my personal favorites:
>
> Point taken.
Actually, I have believed for a very long time that there is a very concrete
way to demonstrate the superiority of the HP3000, IMAGE, and in this
particular case, QueryCalc, over a very similar RDBMS/SQL mix on a UNIX box.
The TPC-D standards, which aren't as commonly quoted or measured as are the
TPC-C, are a set of very standard business reports, representative of the
general kinds of reports that most any business needs to run every day. While
they are specified only for an RDBMS/SQL construct -- and thus would have no
validity outside of that arena -- they can be completely and quite easily
duplicated on an HP3000 using IMAGE and QueryCalc.
Because the particularly odd situation that exists with HP-UX and MPE running
on exactly the same hardware, this is a chance to compare apples to apples --
with NO special modification of any software to the specific circumstances.
I've mentioned this attribute before on this list. While it would be a little
bit of work to set it up, and I'm sure that the results wouldn't be accepted
by the TPC, I am also sure that the QueryCalc/IMAGE combination would "blow
the doors off" (an old drag racing term) of any combination of RDBMS products
that might be run under HP-UX.
Regardless of their general acceptance, these would be concrete numbers.
There would be nothing nebulous about the results.
Wirt Atmar
|
|
|