Hi guys, I want to have a synthetic workload for testing various "performance features" of SQL Server. Here are the requirements of the workload off the top of my head - Scalable. The workload should be able to scale from running on a small desktop level machine (dual core 2 GB RAM with one physical hard disk) to a decent commodity server (four socket quad core 16 GB 6 physical disk) class machine. Real life like. Should represent a real life scenario. Amazon.com model comes to mind. Reproducible / Predictable. Should reproduce the same results no matter how many times it is executed. For a performance validation, being able to reproduce a scenario is very important. Shippable. The workload should be able to zip into a small download so that anyone can reproduce the workload on his / her machine. Other requirements that are good to have but not necessary a deal breaker are - Able to execute from 2 machine set up. I would prefer not to have a application server layer just to keep things simple. Ability to capture performance metrics. Sure one can keep track of individual tests and the performance data using text files, word documents. But it would nice to have a system to keep track of all the tests and corresponding results. Ability to run on other commercial database systems. I am not targeting any other database system than SQL Sever 2005 and future versions of SQL Server. But if I can can have a database agnostic workload, that would be sweet. What do you guys think the requirements of this workload should be? Anything else comes to mind? BTW, Quest Software has Benchmark Factory for databases bit it will not allow me to use more than 20 users after the trial period has expired. Do you guys know of any other product without such limitations?