云计算环境下基于蚁群优化算法的资源调度策略  

Resources Scheduling Strategy Based on Ant Colony Optimization Algorithms in Cloud Computing

作　　者：刘永[1] 王新华[1,2] 邢长明[3] 王硕[1] 

机构地区：[1]山东师范大学信息科学与工程学院,山东济南250014 [2]山东省分布式计算机软件新技术重点实验室,山东济南250014 [3]山东财政学院继续教育学院,山东济南250014

出　　处：《计算机技术与发展》2011年第9期19-23,27,共6页Computer Technology and Development

基　　金：山东省优秀中青年科学家科研奖励基金(BS2010DX032)

摘　　要：针对当前云计算环境中节点规模巨大,单个节点资源配置较低,寻找有效计算资源效率不高的缺点,文中在Google公司的Map/Reduce框架上提出了两个基于蚁群优化的资源调度策略ACO1和ACO2,并在这两个资源调度策略中引入双向蚂蚁机制。在该双向蚂蚁机制中蚂蚁通过相互交流,能够快速地发现合适的虚拟机资源,从而使得Master节点能够快速地为用户任务分配虚拟机。实验结果表明这两个利用了双向蚂蚁机制的资源调度策略显著减少了为用户任务寻找虚拟机的时间,从而使得用户任务能够更快地获得虚拟机,保证用户作业能够按时完成。It presents two resources scheduling algorithms which are named ACO1 and ACO2 respectively for the cloud computing because of the disadvantage that the scale of nodes is huge,the configuration of nodes is not high and the efficiency of finding nodes is low.The two resources scheduling algorithms are based on ant colony algorithm and Map/Reduce frame which belongs to Google＇s company.And two-way ant mechanism is introduced into the two resources scheduling algorithms.In the mechanism the ants can find the virtual machines which perform the tasks fast by the communication of ants so that the Master node can assign the virtual machines to the tasks fast.The experimental result demonstrates that the time to find virtual machines which perform the tasks by ACO1 and ACO2 reduces observably,which advantage of the two-way ant mechanism so that it reduces the time to assign the virtual machines to the tasks and assures the users＇ job can be completed on time.

关 键 词：云 资源调度 蚁群算法 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]