Dateset of “Review of Software Performance and Architecture Analysis Integration”

Software architecture is the most important determinantto systematically achieve quality attributes in a softwaresystem, including software performance. However there is nocomprehensive understanding of why and how to integratesoftware architecture and performance analysis to guide research.To fill this gap, this paper presents a systematic review of 83studies. We focus on six research questions that provide guidancefor researchers and practitioners to gain an in-depth understandingof this research area. These questions addressed: the study purposes(RQ1), the relevant software development activities (RQ2), thetypical study templates (RQ3), the available instruments forautomating the analysis (RQ4), the evaluation methodology (RQ5),and the limitations and future research directions (RQ6). The mostimportant outcome of our study is an identification of criticalresearch gaps and directions, including: 1) the lack of available toolsand benchmark datasets to support replication, cross-validation andcomparison of studies; 2) the need for architecture and performanceanalysis techniques that handle the challenges with emergingsoftware domains, such as blockchain and cyber-physical systems;3) the lack of consideration of factors such as usability thatimpact the practical adoption of the architecture and performanceanalysis approaches; 4) the need for integrating architecture andperformance analysis in Agile settings; and finally 5) the need forthe adoption of modern ML/AL techniques to efficiently integratearchitecture and performance analysis.

软件架构是系统性地在软件系统中实现质量属性（包括软件性能）的最重要决定因素。然而，目前对于如何以及为何将软件架构与性能分析相结合以指导研究尚无全面的理解。为填补这一空白，本文对83项研究进行了系统综述。我们聚焦于六个研究问题，为研究人员和实践者提供指导，以深入理解该研究领域。这些问题涉及：研究目的（RQ1）、相关的软件开发活动（RQ2）、典型的研究模板（RQ3）、用于自动化分析的可用工具（RQ4）、评估方法（RQ5）以及局限性和未来研究方向（RQ6）。本研究最重要的成果是确定了关键的研究空白和方向，包括：1）缺乏可用的工具和基准数据集以支持研究的复制、交叉验证和比较；2）需要处理新兴软件领域（如区块链和 cyber-physical systems）所面临的挑战的架构和性能分析技术；3）缺乏对如可用性等因素的考虑，这些因素会影响架构和性能分析方法的实际应用；4）需要在敏捷环境中整合架构和性能分析；最后，5）需要采用现代机器学习/人工智能技术以有效地整合架构和性能分析。