云数据中心资源利用率均衡的虚拟机调度算法

发布时间：2018-09-13 07:47

【摘要】：云数据中心有大量的服务器,服务器的各种资源利用率不均衡将会导致该服务器不能进一步部署虚拟机,造成资源浪费.现有的研究主要针对某几种资源(例如CPU、内存)来均衡服务器之间的资源利用率以使得云数据中心负载均衡,但很多研究工作没有充分考虑服务器上各种资源之间的利用率均衡问题(例如:服务器上CPU和内存的利用率相差较大),基于此本文提出一种基于资源利用率均衡的虚拟机调度模型VMSM-RB.该模型将虚拟机调度问题抽象为不定维向量装箱模型,其中向量的每一维代表一种类型的资源,本文进一步提出一种自适应粒子群优化算法SA-PSO求解该模型,最后通过仿真实验验证了该模型的有效性.实验结果表明,VMSM-RB模型能够对云数据中心中的虚拟资源进行高效的调度与管理,相比其他同类模型和算法能够更好的均衡服务器上各种资源的利用率,减少资源浪费.
[Abstract]:There are a large number of servers in the cloud data center. The unbalanced utilization of various resources of the server will lead to the server can not further deploy the virtual machine, resulting in waste of resources. The existing research focuses on several kinds of resources (such as CPU, memory) to balance the resource utilization between servers to balance the load of cloud data center. However, many researches have not fully considered the utilization balance between various resources on the server (for example, the difference between the utilization ratio of CPU and memory on the server is quite large). Based on this, this paper proposes a kind of equilibrium based on resource utilization. Virtual machine scheduling model VMSM-RB. The model abstracts the virtual machine scheduling problem into an indefinite dimensional vector packing model, in which each dimension of the vector represents a type of resource. In this paper, an adaptive particle swarm optimization algorithm (SA-PSO) is proposed to solve the model. Finally, the effectiveness of the model is verified by simulation experiments. The experimental results show that the VMSM-RB model can efficiently schedule and manage virtual resources in cloud data centers. Compared with other similar models and algorithms, the VMSM-RB model can better balance the utilization of all kinds of resources on the server and reduce the waste of resources.
【作者单位】：南京航空航天大学计算机科学与技术学院;软件新技术与产业化协同创新中心;
【基金】：国家自然科学基金青年科学基金项目(61202351)资助江苏省普通高校研究生科研创新计划项目(CXZZ13_0171)资助
【分类号】：TP302;TP308

【相似文献】