弹性伸缩技术可以提高用户满意度,很好地解决了资源利用率和应用系统之间的矛盾,是云计算的关键技术之一.目前主流的自动伸缩技术是基于阈值的,但此策略遭遇暂时的负载增加时容易造成频繁伸缩,对系统产生额外耗费以及资源浪费.针对此问题,论文提出一种基于预测的云计算弹性伸缩策略,介绍了三种预测模型.最后根据仿真结果,展示了此策略的优越性.%Elastic retractable teachnology is one of crucial teachnology in cloud computing,which can improve user satisfac?tion and solve the contradiction between resource utilization and application system. At present,the mainstream elastic retractable technology is based on threshold. However,this method always meets frequent stretching when there is temporary load increase, which generate additional costs and waste of resources. Aim at this problem,an elastic retractable strategy based on predictive mod?el is proposed. Three prediction models are introduced. Ultimately,the superior performance of our strategy is confirmed by simula?tion results.
展开▼