Enhancing Microservice Performance: A Hybrid Approach Using Caching and Batching Techniques

Abstract

Nowadays, microservices play a major role in industries by enabling modular and scalable software design through inter-service communication. However, traditional communication methods—typically involving JSON-based API calls—can lead to high latency, increased network traffic, and redundant database queries. These inefficiencies become more prominent when services repeatedly request the same data or make multiple small API calls rather than batched requests. To address these issues, this paper explores two key optimization techniques: caching and batching, along with their combined effect. The objective is to analyze how these techniques individually and together improve performance in microservice-based systems. We conducted experiments under four scenarios: no optimization, caching only, batching only, and caching combined with batching. The implementation was carried out using Spring Boot, Redis for caching, and JPA for batch data access. The results demonstrate that caching alone reduces response time by up to 98.31%, batching alone by approximately 86.97%, and the hybrid approach of caching with batching offers consistent low latency with minimal database load. This confirms that a combined caching and batching strategy provides the most efficient and scalable solution for optimizing inter-service communication in microservices.