Model and Algorithm for Data Distribution in the Microservice Architecture of a Web Application

This paper is devoted to the study and development of a model and algorithm for data distribution in the microservice architecture of a web application, considering the possible inoperability of some services. The research focuses on enhancing the performance of load balancers by considering key parameters such as request intensity, system capacity, and reliability. A mathematical model is formulated using queuing theory and Erlang's formulas to define the optimal function for data distribution. The study proposes an improved algorithm for dynamic load balancing that actively monitors microservice performance, reducing bottlenecks and increasing processing efficiency. Simulation results demonstrate a significant improvement in the output flow of serviced data messages, increasing throughput by 15.25% compared to conventional load balancing methods. The proposed approach enables efficient utilization of system resources, enhances fault tolerance, and improves request execution speeds in microservice-based web applications. These findings highlight the importance of adaptive data distribution mechanisms in high-load environments, ensuring reliability and scalability for modern banking and financial services applications.