Path: utzoo!attcan!uunet!yale!husc6!hscfvax!pavlov
From: pavlov@hscfvax.harvard.edu (G.Pavlov)
Newsgroups: comp.databases
Subject: Re: Oracle Peformance Simulation
Message-ID: <679@hscfvax.harvard.edu>
Date: 23 Nov 88 06:09:54 GMT
References: <3352@newton.praxis.co.uk> <557@4gl.UUCP> <246@daitc.daitc.mil>
Organization: Health Sciences Computing Facility, Harvard University
Lines: 24

>From article <3352@newton.praxis.co.uk>, by ben@praxis.co.uk (Ben Dillistone):
> I am in the process of trying to create a performance evaluation
> environment for an interactive/batch Oracle system.
>. . .
> The problem is simulating variable numbers of users.

  Whatever method you use will be viewed as suspect.  Hopefully by yourself 
  as well as others.

  Quite a long time ago, we attempted to benchmark systems using an approach
  which may give you some ideas.  We wrote a series of programs requiring
  "user interaction" which ran on the machines to be tested and a Compaq por-
  table with software that simply provided the appropriate "response" to the
  programs' prompts.  The Compaq portable was used because we had to travel to
  the machines.  It was also a decent match for the machines we were looking
  at, VAX 750's and peers.


  If I were to try to do something like this again, I would obtain a fast pc
  with Unix and a batch of serial ports.  Though "fast" would probably still
  not be fast enough for a transaction processing simulation on a capable
  host machine.

   greg pavlov, fstrf, amherst, ny