Reusable QoS specifications for systematic component-based design

Abstract

For successful and effective software development the ability to predict impact of design decisions in early development stages is crucial. Typically, to provide accurate predictions the models have to include low-level details such as used design patterns (e.g., concurrency design patterns) and underlying middleware platform. These details influence Quality of Service (QoS) metrics, thus are essential for accurate prediction of extra-functional properties such as performance and reliability. Existing approaches do not consider the relation of actual implementations and performance models used for prediction. Furthermore, they neglect the broad variety of implementations and middleware platforms, possible configurations, and varying usage scenarios. To allow more accurate performance predictions, we extend classical performance engineering by automated model refinements based on a library of reusable performance completions.