论文部分内容阅读
随着云计算技术的发展与运用,云计算在资源的效用比、按需服务等方面优势显著;相对于传统的计算资源构建,云计算凭借其安全性保障、高效的弹性计算资源分配能力、简易的硬件要求等特性,能实现面对不同需求时的计算资源快速弹性构建。基于云计算的弹性资源效用比为研究点,以经典的电信接入随机过程模型—泊松过程为基础,采用相关性时间序列模型对资源预测进行研究,并对相似的多类模型进行仿真与分析;最后以Openstack云平台为基础,结合实际需求对模型进行了工程化实现和初步测试。结果表明,该方法为云计算按需使用和资源弹性构建提供了一种可行的方式,在保证服务等级协议(service-level agreement,SLA)的同时,进一步降低云计算平台的运行损耗,提高资源的效用比。
With the development and application of cloud computing technology, cloud computing has significant advantages in terms of resource utility ratio and on-demand service. Compared with traditional computing resource building, cloud computing, with its security guarantee and efficient elastic computing resource allocation capability, Simple hardware requirements and other features, to meet the different needs of the computing resources when the rapid and flexible construction. Based on the cloud computing elastic resource utility ratio as the research point, based on the classical telecommunication access stochastic process model - Poisson process, the correlation time series model is used to study the resource forecasting, and the simulation of similar multi-class models Finally, based on the Openstack cloud platform, the model was implemented and tested according to the actual requirements. The results show that this method provides a feasible way for on-demand use of cloud computing and resource elasticity construction. While guaranteeing the service-level agreement (SLA), this method can further reduce the operation loss of cloud computing platform and improve the resource The utility ratio.