基于模因优化和循环调度的多目标负载均衡技术  

MULTI-OBJECTIVE LOAD BALANCING TECHNOLOGY BASED ON MEMETIC OPTIMIZATION AND CYCLIC SCHEDULING

作　　者：罗宁[1] 李璐 唐忠 Luo Ning;Li Lu;Tang Zhong(Network and Modern Educational Technology Center,Guangxi Economic and Trade Vocational Institute,Nanning 530021,Guangxi,China;School of Information and Management,Guangxi Medical University,Nanning 530021,Guangxi,China)

机构地区：[1]广西经贸职业技术学院网络与现代教育技术中心,广西南宁530021 [2]广西医科大学信息与管理学院,广西南宁530021

出　　处：《计算机应用与软件》2021年第4期69-74,共6页Computer Applications and Software

基　　金：广西壮族自治区高等学校科学技术研究项目(KY2015YB447);广西教育科学“十二五”规划教育信息化工作专项课题(2015ZKY17);广西学位与研究生教育改革课题(JGY2017038)。

摘　　要：针对现有多目标调度方法所需时间较长以及处理突发情况时性能降低的问题,提出一种基于模因优化和循环调度的多目标负载均衡技术。使用突发检测器检测发送到云服务器的用户请求,确定负载状态。基于测器结果,应用不同的负载平衡算法来高效地调度用户任务。利用选定的负载平衡算法将用户请求任务调度到资源最佳的虚拟机上,保证在最低的时间消耗内达到负载均衡的状态。实验结果表明,与其他算法相比,该方法在多个性能指标上具有明显优势,可以提高调度效率的同时,最大限度地降低云中的能源使用。Aiming at the problem that the existing multi-objective scheduling method requires long processing time and performance degradation when dealing with burstiness workloads,a multi-objective load balancing technology based on memetic optimization and cyclic scheduling is proposed.It used a burst detector to detect user requests sent to the cloud server to determine the load status.Then based on the detector results,different load balancing algorithms were applied to efficiently schedule user tasks.Finally,the selected load balancing algorithm was used to schedule user requests to the best-resourced virtual machine to ensure load balancing within the minimum time consumption.The experimental results show that,compared with other algorithms,the proposed method has obvious advantages in multiple performance indicators,which can improve the scheduling efficiency and reduce the energy use in the cloud to the greatest extent.

关 键 词：云计算 模因优化 循环调度 多目标负载均衡 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]