Optimizing Microservices Deployment with Kubernetes: A Performance and Scalability Analysis

This thesis explores the design, development, and deployment of a federated authentication mechanism for microservices, specifically tailored for the PartsCoder Middleware system at Codebeex SRL. As organizations increasingly adopt microservices architectures, ensuring seamless and secure authentication across distributed services becomes crucial. This project investigated various federated authentication technologies to enable secure, efficient access control and user identity verification across multiple microservices without compromising system performance. Through this study, I developed Python-based FastAPI microservices with a JWT (JSON Web Token)-based authentication model to achieve secure user verification across independent service endpoints. The JWT mechanism allows decentralized authentication across the PartsCoder application, minimizing risks associated with single-point failures and enhancing scalability and reliability. Key microservices for survey management in the PartsCoder Middleware were identified, designed, and implemented, including secure data exchange protocols to safeguard user information.The deployment of these microservices leveraged Docker and Kubernetes, enabling efficient scaling and management of containers across clusters. Kubernetes facilitated robust orchestration and fault tolerance, streamlining the integration of the new authentication framework into the existing infrastructure at Codebeex SRL. This thesis demonstrates how the federated authentication model enhances both security and performance in microservices-based applications, providing insights and best practices for authentication across similar distributed systems in production environments.