Orchestration of service-oriented applications with reactive programming techniques

Abstract

The object of research is the modular approach to application development using SOA, as well as the comparison of synchronous and asynchronous request processing methodologies using a reactive programming architecture. SOA allows applications to be divided into independent components, ensuring easy integration and scalability in distributed computing environments. With SOA, it is possible to create a network of loosely coupled services, providing users with the flexibility to develop applications tailored to specific needs.

One of the main issues is thread blocking and system instability under heavy loads when using synchronous methods. The study compares synchronous and asynchronous request processing methodologies using WebFlux, and examines key components of SOA, such as service discovery mechanisms and interaction models, particularly orchestration and choreography.

The results show that asynchronous approaches, using a non-blocking, event-driven architecture, reduce the number of active threads, increase system resilience, and improve performance. This is because the proposed non-blocking, event-driven approach has several features, including reducing thread blocking and enhancing system stability under heavy loads. Synchronous methods, while straightforward, have drawbacks such as thread blocking and system instability under excessive loads.

As a result, there is a high efficiency in processing a large number of requests in real-time. Compared to similar known approaches, this provides advantages such as increased system resilience and efficient resource utilization, making this approach particularly useful for scalable application architectures in distributed computing environments.