Modeling System Performance with Production Data

When creating system performance models, the primary challenges are where and how to start. Whatever the performance characteristics being estimated or modeled, we need a solid approach that addresses both business and system needs. All too often performance tests inadvertently mix load and stress scenarios with little regard for how this will confound recommendations and business decisions. If you are a test manager, a business process owner, or you simply want to better understand performance testing, you will be interested in William Hurley’s case studies. Will presents two real-world examples that demonstrate the impact on business decisions, and show how to use production data and statistical modeling to improve both the analysis and business decisions. The first study is a cloud-based performance testing engagement. The second is a back-end re-hosting study where acceptance criteria was based on an achieving “equal or faster” performance. Take away new insights and approaches to improve performance-based decisions for your organization.