Diseño de una arquitectura en microservicios con indexación dinámica y distribuida para optimizar la búsqueda en tiempo real de la disponibilidad en un sistema de reservas empresarial coexistente con una arquitectura monolítica

Abstract

This project presents the design of a microservices-based architecture with dynamic and distributed indexing, aimed at optimizing real-time availability search in an enterprise reservation system that coexists with a monolithic architecture. To address the limitations of scalability and performance identified in the latter, the study proposes a technical solution through the progressive adoption of the Strangler pattern, integrating specialized technologies such as Elasticsearch for distributed indexing, Redis for caching, and RabbitMQ for asynchronous data synchronization. Experimental results showed significant improvements in latency and processing capacity. Under high-traffic conditions, the original monolithic version recorded average response times exceeding 30 seconds. With the integration of Elasticsearch, this time was reduced to under 100 milliseconds in the optimized system, indicating a drastic improvement in query efficiency. Additionally, Redis helped stabilize response times for recurring queries in high-concurrency scenarios. The technical validation demonstrated that the designed architecture not only meets scalability and performance requirements but also lays the foundation for the sustainable evolution of the monolithic system, enabled by the intrinsic modularity of microservices. The research provided valuable insights into how agile methodologies and the strategic implementation of microservices can transform legacy enterprise architectures, fostering more flexible, efficient, and adaptive technological environments.