Re: TCP benchmarks

From: tolly brumbelow <tolly.brumbelow_at_bridge.bst.bls.com>
Date: 1996/04/15
Message-ID: <4ktgiv$pqi_at_atglab10.atglab.bls.com>#1/1


Snorri Bergmann <snorri_at_strengur.is> wrote:

>HiHo!
 

>I just got Computerworld from Oct. 30th and saw two ads
>from Oracle (pages 7 & 9). I'd like to hear peoples opinion
>on these facts:
 

>1) All O benchmarks are "unavailable" until march 1996 and are made
>using an unreleased product, Oracle 7.3.
 

>2) O compares results to Sybase 10 only, although Sybase 11 benchmarks
>are available at the time the ad is made. (SUN 2000E 4544 TPC/$396/Tpc).
>They are however "unavailable" just as O 7.3, but I guess mine
>"unavailable" are better than yours.
 

>3) O compares (on page 9) it's HP K410 benchmark to Informix's T500
>benchmark which was WITHDRAWN at the time Oracle submitted the K410 one.
>Oracle has done a benchmark on T500 w/12 CPUS at a similar $/TPC as
>Informix (but more cpus, faster fe's, newer HP-UX version and absolutely
>less expensive hardware because the O benchmark is like 9 months newer
>than the Informix one.)
>The Informix/T500 benchmark is over a
>year old (since Q4 '94), and in fact most of the benchmarks on p.7
>(apart from the Oracle ones) are "old". Maybe the competition should
>dig up old tcp-b benchmarks Oracle has withdrawn and use them in
>comparison.
 

>I find it highly unlikely that ads like these are in the spirit of
>the TCP-Council. Informix and Sybase could undoubtly create similar ads
>by just ignoring/neglecting some competitors' benchmarks and in the
>end the customer would be confused as hell.
 

>Take care,
 

>-Snorri

I have heard oracle has pulled the ad due to lies in the ad about the informix benchmark. Not only is it an old benchmark but informix was also on a k400 series not a t500 and both had 4 cpus but the oracle ad says informix was on a bigger box and in real life informix benchmark was on a smaller box.

-- 
Tolly Brumbelow, Informix DBA
email:tolly.brumbelow_at_bridge.bellsouth.com
Received on Mon Apr 15 1996 - 00:00:00 CEST

Original text of this message