Towards a Performance Testing Body of Knowledge (PTBOK)

• Ano de defesa: 2021
• Autor(a) principal: Martins, Guilherme Legramante
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal do Pampa; UNIPAMPA; Mestrado Profissional em Engenharia de Software; Brasil; Campus Alegrete
• Palavras-chave em Português: Engenharia de software; Software - desempenho; Software - Testes; Software Engineering; Computer software - Testing; Software - performance; CNPQ::CIENCIAS EXATAS E DA TERRA
• Link de acesso: http://dspace.unipampa.edu.br:8080/jspui/handle/riu/5919

Due to the growing market demand, web applications need to quickly respond to users’ requests, since their engagement may be inclined to fluidity and agility in interactions. For this reason, through performance testing, we may map out the scalability of the application and identify bottlenecks that may affect its performance. There are some studies in academia and industry that provide information for conducting this type of test. However, there is no standardization of procedures, and the information is not organized in a way that facilitates the performance test execution in its complete cycle, from its conception to the generation of the reports. Hence, our research seeks to identify the main inputs, outputs, and processing involved in the performance testing, so that we developed the Performance Testing Body of Knowledge (PTBOK). This body of knowledge aims to assist the preparation and conduction of the performance test. Then, we did a Systematic Literature Review (SLR) that, mapped out the performance testing throughout its life cycle. We found and detailed thirty seven (37) papers, elaborating a feature model from them. We also carried out another empirical study: a survey, which sought to identify (as well as the SLR) data related to the performance testing process from the software industry. After conducting these two studies, we filtered and merged the results, we started the PTBOK creation. For the modeling of the process, we chose the Software & Systems Process Engineering Metamodel (SPEM), as we understand that it is a viable and adequate alternative since it is specifically for modeling software processes. PTBOK is already available and we evaluated it through a survey of performance testing experts.