云计算中虚拟机负载均衡调度方法研究及优化方案

Load balancing in cloud computing is essential for optimizing resource utilization and improving system performance. This study focuses on the research and development of virtual machine scheduling methods for load balancing in cloud computing. The main objective is to evenly distribute tasks among virtual machines, taking into consideration factors such as affinity, genetic algorithms, and migration cost. The study first analyzes the distribution of tasks and calculates the values of vc(j, T) and vd(j, T), which represent the load and the distance between virtual machines and tasks, respectively. Then, a genetic algorithm is employed to obtain the final solution space S and identify the mapping scheme S′ with the minimum cost factor. A comparison is made between S and the optimal mapping scheme S′, and based on network congestion conditions, virtual machines are migrated to corresponding hosts to achieve load balancing. Experimental results indicate that the proposed virtual machine scheduling algorithm outperforms the first-fit and round-robin algorithms, as well as the NABM algorithm, in terms of system load balancing and migration costs. The performance improvements are significant across various metrics, demonstrating the comprehensive advantages of the proposed approach. In conclusion, this research contributes to the advancement of load balancing techniques in cloud computing, particularly in the context of virtual machine scheduling and migration. The use of affinity and genetic algorithms in conjunction with effective load balancing strategies has shown promising results in improving system performance and resource utilization. This study serves as a valuable reference for researchers and practitioners in the field of cloud computing and distributed systems. Keywords: Cloud computing; Virtual machine scheduling and migration; Affinity; Genetic algorithm; Load balancing.