Work

A Containerized Load-Balanced Microservice Scheduling Framework for Execution Time and Deployment Cost Optimization in Cloud Environments

Year: 2026
Type: article
Abstract: Cloud computing has seen widespread adoption of containerized microservice architectures, largely owing to their scalability, deployment efficiency, and operational flexibility. Despite this growth, prevailing microservice scheduling methods continue to struggle with challenges such as imbalanced workload distribution, elevated deployment costs, server overload, network bottlenecks, and excessive execution latency, all of which compromise Quality of Service (QoS) and diminish resource utilization efficiency.To address these limitations, this research introduces the Containerized Microservice Load Balancing (CMLB) framework, designed to optimize microservice scheduling within cloud environments. The framework combines distributed resource management with a hierarchical load balancing architecture composed of a Master Load Balancer (MLB) and multiple Local Load Balancers (LLBs) operating across dedicated Microservice Controllers (MSCs). Complementing this architecture, a new CMLB scheduling algorithm is proposed to simultaneously improve deployment cost-efficiency, execution speed, resource allocation, and system scalability. The underlying optimization problem is cast as a multi-objective scheduling model that seeks to minimize both application deployment cost and microservice execution time, while respecting resource capacity and workload constraints. To validate the framework, experiments were carried out using Google Cluster Trace datasets within a Docker-based microservice environment built on Spring Boot, Spring Cloud, Netflix, and Eureka service discovery. The CMLB algorithm was benchmarked against established scheduling approaches, including EPTA, Spread, Binpack, Random, and Optimal-VM. The results show that the proposed framework markedly outperforms existing methods in terms of workload balance, deployment cost reduction, execution time, traffic spike mitigation, and server overload prevention. These findings affirm that the CMLB framework delivers meaningful gains in scalability
Language: English
Open Access status: diamond
OpenAlex ID: W7218093269
FWCI:
Cites:
Cited by:
Related to: 10