Cloud Load Balancing Algorithms Performance Evaluation Using a Unified Testing Platform

Due to its improved response time, availability, and efficiency, load balancing emerged as an essential framework for designing high-performance distributed computing systems. This paper introduces a unified testing platform that objectively compares load balancing algorithms and measures their performance. It employs various request patterns and load types to simulate real-world conditions. We evaluate a selection of static and dynamic algorithms on throughput, response time, and failure rate metrics. The results show that most, but not all, dynamic algorithms perform better than static ones.