摘要
针对云平台下资源利用率不高,资源变化的准确预测缺乏等问题,设计了基于混合预测的云平台资源分配方法。该方法结合服务资源需求的周期性特点,通过快速傅里叶变换方式对服务资源需求的周期性进行判断,利用马尔科夫过程预测缺乏周期性的资源请求,再根据预测结果自适应地分配虚拟机资源。实验结果表明,该方法可以对服务资源需求进行准确预测,合理分配虚拟机资源,改善虚拟机资源的利用率,并且能够减少服务等级协议的违反次数。
According to the low cloud platform resource utilization rate, and the lack of accurate prediction of the variation of resources, a cloud platform resource allocation method based on hybrid prediction is designed. The method combines the cyclical nature of the service resource requirements with the fast Fourier transform（FFT） method to judge the periodicity of service resource demand. Markov process is used to predict the lack of periodic resource requests and then the virtual machine resources are adaptively allocated according to the prediction results. Experimental results show that the method can accurately forecast service resource requirements of the rational allocation of virtual machine resources and improve the utilization of virtual machine resources, and reduce the number of service-level agreement（SLA） violations.
作者
杨娜
李红娟
YANG Na;LI Hong-juan(School of Information Engineering;School of Software,Henan University of Animal Husbandry and Economy,Zhengzhou 450044,China)
出处
《控制工程》
CSCD
北大核心
2018年第11期2102-2108,共7页
Control Engineering of China
基金
郑州市社会科学重点调研课题(JX20150044)
关键词
虚拟机
资源分配
快速傅里叶变换
马尔科夫链
Virtual machine
resource allocation
fast Fourier transform
Markov chain
