摘要
文章在遗传算法的基础上,构建了一种用于虚拟机簇部署的调度算法。此算法的核心是适应度函数的构建,构建的基础是2个优化目标:一个是结合CPU、内存、硬盘、带宽等参数的资源利用率,另一个是基于计算时间的处理效率。为了考察此算法的效果,选取启发式算法和GC(graph cut)算法作为比较算法。实验结果表明基于双目标优化的遗传迭代调度算法,具有更高的资源利用率和处理效率。
Based on the genetic algorithm, a scheduling algorithm for the virtual machine cluster was developed. The core of this algorithm is the construction of fitness function, and its basis is two opti- mization objectives： one is resource utility with four parameters of CPU, RAM, hardware and band- width; the other is processing efficiency based on computing time. In order to investigate the effect of this algorithm, heuristic algorithm and graph cut（GC） algorithm are selected as the comparison algo- rithm. The experimental results show that the genetic iterative scheduling algorithm based on double objective optimization has higher resource utility and processing efficiency.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
北大核心
2016年第5期632-635,共4页
Journal of Hefei University of Technology：Natural Science
基金
广东省省级科技计划资助项目(2014A010103002)
东莞市高等院校
科研机构科技计划一般资助项目(2014106101037
2014106101033)
东莞职业技术学院政校行企合作开展科研与服务资助项目(ZXHQ2014d010)
关键词
虚拟机簇
遗传算法
资源利用率
处理效率
virtual machine cluster
genetic algorithm
resource utility
processing efficiency
