Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!usc!apple!uokmax!munnari.oz.au!metro!cs.uow.edu.au!ph
From: ph@cs.uow.edu.au (Rev Phil Herring, DD (Ret.))
Newsgroups: comp.databases
Subject: Re: Why is Oracle better than Ingres
Message-ID: <1990Dec7.034816.3234@cs.uow.edu.au>
Date: 7 Dec 90 03:48:16 GMT
References: <734@keele.keele.ac.uk> <2060006@hpcuhc.cup.hp.com>
Reply-To: ph@wyvern.cs.uow.edu.au (Rev Phil Herring, DD (Ret.))
Organization: The University of Wollongong, NSW, Australia
Lines: 26

In article <2060006@hpcuhc.cup.hp.com> dhepner@hpcuhc.cup.hp.com (Dan Hepner) 
  writes:
>>ORACLE was over ten times faster than Ingres in that case.
>
>All major DBMS vendors, including Oracle, are regularly plagued
>by such reports.

I *did* point out the circumstances of the test (very high insert
rate). I also pointed out that this was a rather odd example.

>The default assumption, as supported by many cases, is that
>the slow system can be made to perform at something like the
>speed of the fast one. This may take some cleverness, or
>may take some configuration modification not well explained
>by the vendor, but it can almost always be done.

The vendors were all involved. They were given specs and wrote
their own benchmark programs. The only exceptions were Oracle
and Informix - I wrote their test programs. I can only assume, 
therefore, that the Ingres people knew how to make their system
run as fast as it could, and that they did their best. By the way, 
I didn't try to tweak ORACLE to the limit. The database was 
pretty much "out of the box".


-- Phil.


Brought to you by Super Global Mega Corp .com