Saxena, D.Bhowmik, B.2026-02-062024VLSI SATA 2024 - 4th IEEE International Conference on VLSI Systems, Architecture, Technology and Applications, 2024, Vol., , p. -https://doi.org/10.1109/VLSISATA61709.2024.10560139https://idr.nitk.ac.in/handle/123456789/29029Microservice architecture has become a widely accepted solution to address the challenges, particularly scala-bility, deployment, and flexibility associated with monolithic architecture. A vital attribute of the microservices architecture is its capability to handle load balancing on a large scale. The load balancer collaborates with a scaler to distribute the workload efficiently across multiple instances. In the literature, different studies employ load-balancing algorithms for efficient microservice load balancing. These works overlook cloud-based microservice applications or focus solely on virtual machines, neglecting containers. This paper addresses these limitations by comparatively assessing selected load-balancing algorithms. The three most used algorithms, random, round-robin, and least connection, are studied on a microservice application. The extensive experiments are conducted using Elastic Container Service (ECS) of Amazon Web Service (AWS) for containerized cloud setup where each service resides in a cluster and traffic is generated through Locust. Experimental results show that throughput and response time range of 6.2-288.7 and 312.2-3375.8 ms, respectively. © 2024 IEEE.Cloud-based MicroserviceContainersLoad Bal-ancingPerformance MetricsScalingVirtual Machine