| The DeWitt-Wisconsin benchmark was developed to establish the need
for database machines. It is geared to producing results that force
the database machine to be considered. Normally, Oracle and Ingres
are the big "pushers" of that benchmark. If I remember correctly,
it is not very weighted for update activity. More emphasis is on
querying, lots of selects...
---- Michael Booth
|
| BITTON, D., DEWITT, D.J. AND C. TURBYFIL, "Benchmarking Database
Systems: A Systematic Approach," Computer Sciences Department Technical
Report #526, Computer Sciences Department, University of Wisconsin,
December 1983. This is a revised and expanded version of the paper
that appeared under the same title in the Proceedings of the 1983
Very Large Database Conference, October, 1983.
Boral, Haran and DeWitt, David J., "A Methodology for Database System
Performance Evaluation," Computer Sciences Dept. Tech. Report #532.
This research was partially supported by the National Science
Foundation under grant MCS82-01870 and the Department of Energy
under contract #DE-AC02-81ER10920.
Basically, the first report explains the reasons for using a
a simple, stand-alone database system to begin benchmark developements.
The second article expands on that study into a mutli-system
environment. Also, the second article highlights the key techniques
and strategies that ought to be covered in the study.
I'll dig up the article and write an article report.
|