| Design patterns, components, and frameworks serve as a platform for building various distributed systems; they enable software reuse and save development time. In particular, communication framework design patterns are widely used to develop distributed systems including high-performance servers, telecommunication systems, and control systems. Unlike components, design patterns are often descriptions of a programming approach and need to be reified for each application.;Many design patterns that are used in communication frameworks have a significant impact on the overall performance and scalability of the system in terms of throughput and latency. Therefore, the choice of design patterns and their configuration can impact the overall performance an application.;This objective of our research is not only to allow software developers to predict the performance of their frameworks at design-time, but also to provide reusable models for software performance of many commonly used design patterns. We explore the idea of composability of design patterns by illustrating how some design patterns may be combined to build communication frameworks.;Currently, there is no general way to quantify the performance of components and design pattern implementations across various dimensions such as throughput and latency. Toward this goal, we make the following contributions: (1) we develop analytical models for various design patterns using queuing models, (2) we present a technique to analyze the performance of combinations of design patterns as observed in real-world applications, and (3) we validate the models using empirical measurements. |
