论文部分内容阅读
与网格、集群等传统计算模式相比,云计算的关键优势在于为用户提供了一种利用远程计算资源的实用商业模型.云服务提供商把服务层指标转换为操作层指标,根据服务级别协议动态管理云计算资源.本文研究如何基于服务级别协议在不同的客户之间动态分配云资源池资源,以获得最大收入.借助排队论模型,把资源分配描述为基于价格机制、服务请求到达率、服务速率、可用资源等约束条件的数学极值问题,并通过拉格朗日乘子法给出了最优数学解.分别以合成数据和网络日志数据模拟请求到达率,对理论结论进行了实验验证.实验结果表明本文算法有优于相关工作.
Compared with traditional computing models such as grid and cluster, the key advantage of cloud computing is that it provides users with a practical business model that makes use of remote computing resources. Cloud service providers convert service-level metrics into operational-level metrics, Protocol to dynamically manage cloud computing resources.This paper studies how to dynamically allocate cloud resource pool resources among different clients based on service level agreements to obtain the maximum revenue.With the queuing theory model, resource allocation is described as price mechanism, service request arrival rate , Service rate, available resources and other constraints of the mathematical extreme value, and by Lagrange multiplier method gives the optimal mathematical solution.The synthetic data and network log data simulation request arrival rate, the theoretical conclusions were Experimental results show that our algorithm is superior to the related work.