Sarcouncil Journal of Applied Sciences Aims & Scope

Abstract: Microservices architecture has become a dominant paradigm in modern software development, offering flexibility, scalability, and modularity. However, its distributed nature presents critical challenges in ensuring security, maintaining data consistency, and achieving high performance, particularly in enterprise-grade Java applications. This study aims to address these challenges by proposing an integrated optimization framework that enhances microservices architecture across three key dimensions: security, consistency, and performance. Using Java-based microservices developed with Spring Boot and deployed on a Kubernetes environment, this research implements security mechanisms including OAuth2, JWT, HTTPS, and mutual TLS to reduce system vulnerabilities. Distributed transaction patterns such as the Saga pattern, Event Sourcing, and CQRS were evaluated to improve data consistency across services. High-performance computing was addressed through the adoption of reactive programming models, leveraging asynchronous and non-blocking processing via Spring WebFlux and Project Reactor. The experimental evaluation, supported by statistical analyses such as paired t-tests and ANOVA, revealed significant improvements in all three areas. Security risk scores dropped by over 70%, while CQRS provided the highest transaction success rate (98.1%) and lowest data anomaly rate. Reactive Java significantly reduced response time and improved throughput under heavy load, while also reducing CPU and memory usage. This study concludes that a holistic approach combining secure communication, reliable data handling, and efficient computing can lead to robust and scalable Java-based microservices systems. The findings provide practical insights for developers and system architects designing modern cloud-native applications