Sustaining a Reusable High-Quality Monitoring Framework for Software Engineering Research

点击翻译按钮获取中文摘要

Application-level monitoring and dynamic analysis of software systems are a basis for various tasks in software engineering research, such as performance evaluation and architecture recovery for reverse engineering. The Kieker research software provides monitoring, analysis, and visualization support for these research purposes. It commenced in 2006, and grew toward an open-source software that has been employed in a variety of software engineering research projects over the last decade. Several research groups constitute the open-source community to advance the Kieker software framework. With SustainKieker, we intend to sustain Kieker as a reusable high-quality monitoring framework that will be employed in a wider variety of research areas and used by a larger community. Extended interoperability is a goal for SustainKieker. To improve the usability and the initial skill training for Kieker, we will also build tutorial demos for the new research areas. One goal is to integrate the user guides with the online demo applications to result in interactive tutorials. Extended automated quality assurance is a goal for SustainKieker. Specifically, we plan to significantly increase the test coverage, introduce static architecture checks, extend regression benchmarking, and adopt GitOps for state-of-the-art automation. This way, SustainKieker may also serve as a blueprint for other research software projects that strive for automating their quality assurance. With SustainKieker, we plan to create better conditions for further development via remodularizing Kieker’s software architecture. The GitOps procedures will also improve the process for community contributions. With SustainKieker, we also contribute to the seminal work on research software engineering via software engineering research toward establishing research software science as a discipline. We will address this goal via empirical software engineering research. As specific measures, we will empirically evaluate the effect of tutorials with embedded online demos and the impact of hackathons. We are well aware that controlled experiments with students and action research with hackathons has limited external validity. External validity is the extent to which you can generalize the findings of a study to other situations. However, we consider this to be an initial viable and feasible step toward research software science within this proposed project.