Akademik Veri Yönetim Sistemi
2025 Latin American Conference on Computational Intelligence-LA-CCI-Annual, Meksika, 5 - 07 Kasım 2025, ss.13-18, (Tam Metin Bildiri)
Microservice architectures utilize diverse caching methods to improve the performance of web applications. This key design feature plays a crucial role in maximizing the efficiency of services within distributed architectures. In this paper, we conduct an empirical analysis and discussion of two caching patterns: the Embedded Pattern and the Embedded Distributed Pattern. The performance of both patterns that use Hazelcast caching technology under different loads was analyzed on a testbed developed in a microservice architecture. The test results provide valuable insights regarding the optimal conditions for applying each caching pattern. Experimental results demonstrated that the number of instances in the Embedded Distributed Pattern improves the performance of the ecosystem when handling more than 1000 user requests while the approach requires additional resource utilization. We also plan to consider these insights in the development of an AI-driven aquaponics system as part of our newly funded research project.