云环境下的绿色服务组合方法研究

发布时间：2019-04-27 16:58

【摘要】：近年来随着Web服务技术的快速发展和云计算的逐渐普及,越来越多的大型IT公司都在为中小企业和个人开发者提供云服务。随着云服务的普及和应用,服务选择和组合方法成为重要的研究热点。同时,由于大量网络服务部署在云数据中心上,服务的选择不可避免地将受数据中心底层性能的影响,进而影响服务组合性能。然而,在传统云计算Web服务组合技术中,大多数现有的服务组合方案,只在用户角度专注于寻找聚合QoS值的最优值(如响应时间,吞吐量等)的具体的服务,但未考虑服务组合的成本和依赖于云数据中心拓扑结构和底层参数的能耗和网络资源消耗。对于云计算基础设施服务提供商来说,能耗和网络资源消耗直接关系到运营成本,通过降低资源消耗可以提高云计算基础设施服务提供商的经济效益,同时也与国家的节能减排政策相吻合。本文对服务组合目标中的QoS效用、电力消耗和网络资源消耗之间的权衡进行了研究并进一步提出了云环境下的绿色服务组合方法,该方法引入了电力消耗模型和网络资源消耗模型,旨在最大化QoS效用的同时降低云数据中心的电力消耗和网络资源消耗。为了对该方法进行评估,本文对开源云计算仿真引擎CloudSim进行扩展并对该方法和传统服务组合方法进行对比仿真,然后对仿真结果进行了参数分析和结果讨论,验证了该方法对于云数据中心能耗和网络资源消耗控制的有效性。论文最后总结了相关工作,并提出了下一步的研究方向。
[Abstract]:In recent years, with the rapid development of Web services technology and the gradual popularization of cloud computing, more and more large-scale IT companies are providing cloud services for small and medium-sized enterprises and individual developers. With the popularity and application of cloud services, service selection and composition methods have become an important research focus. At the same time, because a large number of network services are deployed in the cloud data center, the choice of services will inevitably be affected by the underlying performance of the data center, and then the performance of the service composition will be affected. However, in traditional cloud computing Web service composition technologies, most existing service composition schemes focus only on finding specific services with the best aggregate QoS values (such as response time, throughput, etc.) from the user's point of view. However, the cost of the service portfolio and the energy consumption and network resource consumption that depend on the cloud data center topology and underlying parameters are not considered. For cloud computing infrastructure service providers, energy consumption and network resource consumption are directly related to operational costs, and reduce resource consumption can increase the economic benefits of cloud infrastructure service providers. At the same time, it is also consistent with the national energy saving and emission reduction policies. In this paper, we study the tradeoff between QoS utility, power consumption and network resource consumption in service composition goals, and propose a green service composition method in cloud environment. This method introduces the power consumption model and the network resource consumption model to maximize the utility of QoS while reducing the power consumption and network resource consumption of cloud data centers. In order to evaluate this method, the open source cloud computing simulation engine CloudSim is extended and compared with the traditional service composition method. Then the simulation results are analyzed and discussed. The effectiveness of the proposed method in the control of energy consumption and network resource consumption in cloud data centers is verified. Finally, the related work is summarized, and the future research direction is put forward.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.09

【相似文献】