Automating the Assessment of the Performance Quality Attribute for Evolving Software Systems: An Exploratory Study

This paper describes an exploratory study for the evaluation of the performance quality attribute for releases of the same system. The main aim is to reveal performance degradations of architectural scenarios and their possible causes. Three software systems from different domains are used in our study, including a large-scale web system (SIGAA), a UML modeling tool (Argo UML), and a client-server framework for development of network applications (Netty). The data collection of the study is accomplished using a scenario-based approach that uses dynamic analysis and code repository mining to provide an automated way to reveal degradations of scenarios on releases of software systems. The results of our study show the feasibility of the approach to determine the causes of the performance degradations of scenarios, including the degraded and changed methods of scenarios, and the issues that have affected them.