db_STRESS Benchmark - http://dimitrik.free.fr/db_STRESS.html
This benchmark has a long history and based on a true customer's workload. My goal here is to not only measure a highest TPS value for a given database, but also follow its stability curve, response time evolution, and scalability with growing workload. I want to know where is the critical point of the database server and how it performs on even overloaded system...

Интересное нагрузочное тетстирование СУБД
Oracle v9+ (and higher)
- on Solaris/SPARC for the moment only
PostgreSQL v8+ (and higher)
- Solaris/SPARC, Solaris/x86, Linux/x86
MySQL v3+ (and higher)
- Solaris/SPARC, Solaris/x86, Linux/x86

Неплохо бы его адаптировать под Информикс, (да еще и на Винь :)
Попробую пообщаться с автором на этот счет.

добавляю в соответствующий топик:

Writing a Simple Informix Engine Benchmark in Python
Posted by Andrew Ford @ 09:32 AM ET | Apr 2, 2008
http://www.ibmdatabasemag.com/blog/main/archives/2008/04/writing_a_simpl.html

Writing a simple benchmark in python for your environment is easy and a good way to get a feel for the maximum transaction rates of your platform. If you know this information you will be able to identify the limits of your system and prepare for growth before you hit those limits. I want to show you how I wrote a script to benchmark for the account lookup and balance update for our stored value telecom application using threads to simulate our multiple front ends accessing the account database...