Mapeamento de dependências não intrusivo em tempo real para microsserviços no Kubernetes

• Ano de defesa: 2025
• Autor(a) principal: Silva, Alcides Mignoso e
• Orientador(a): Guardia, Hélio Crestana
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Inglês: Microservices architecture; Service dependencies; Non-intrusive automatic mapping
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO
• Link de acesso: https://hdl.handle.net/20.500.14289/22549

The ever-evolving landscape of microservice architecture demands innovative solutions for improved performance and environment management, especially under varying request loads. While strategies developed for managing monolithic systems have been effective in the past, they may not be effective in dealing with the challenges of the microservice-based applications, which often present a high level of interdependency between services. Considering the potential benefits of knowing the relationships among microservices and understanding their behaviors, this work introduces a modular and adaptive tool that maps the microservice dependencies within systems and use the data to provide insightful information about the services’ communications patterns. The central hypothesis this study posits is that by using existing network tools, such as Cilium, it is possible to automatically map service dependencies and communication ratios in a microservices environment and use the connections information to generate dependency data that can serve as input for application management. Given its versatility and large adoption, Kubernetes is the chosen development platform, enhancing the scope and usability of the results obtained with this research. Cilium and its eBPF-based technology are used for collecting the information needed for automatically creating the envisioned microservices communication maps, and authoring applications are introduced to transform and process the collected information. The results of this research may benefit practitioners seeking to optimize the architecture of their microservice-based applications, and also contribute to the broader field of application development by emphasizing the importance of considering services dependencies and characteristics in services management. Thus, the findings shall inspire future research endeavors to explore innovative approaches in addressing the unique challenges posed by microservices architecture by utilizing the intrinsic communication information between services.