基于OpenStack的业务云平台资源分配模型的研究与实现
发布时间：2019-05-31 17:47
【摘要】：云计算技术的出现引发了IT领域的重大变革,它不仅改变了IT公司交付软件和硬件产品的模式,也影响了用户通过互联网存储和获取信息的方式。云计算技术凭借其自身巨大的优势,吸引了众多的服务提供商将应用服务部署到云上。在业务云平台上,资源分配的效率将直接影响到服务运行的性能和成本,是业务云平台商业化过程中必须解决的一个关键问题。因此,研究在业务云平台上如何进行高效的资源分配具有重要的研究意义和应用价值。 为此,我们学习和了解了与云计算相关的基础知识和关键技术,并比较了当前几种常见的开源IaaS云平台。根据各个平台的特点及本研究的实际需求,选取了OpenStack开源项目用于构建私有业务云平台,并将多媒体分享业务部署到该平台上。在学习了云计算环境中几种常见的分配算法后,本论文对业务云平台上的资源分配问题进行了分析和建模,根据业务云平台的特点提出了一个资源分配模型,并将其实现。分配模型分为两级：物理机级别和虚拟机级别。在物理机级别,选用OpenStack项目的简单分配算法为各个虚拟实例分配平台中的物理资源；在虚拟机级别,将问题看作是尺寸可变的装箱问题,提出一种算法来实现以服务组件为单位的资源分配。 经实验验证,提出的资源分配模型在业务云平台上实现了预期的分配效果。分配结果可以保证满足服务性能的要求,同时还可以减少一些额外开销,使资源的利用更加有效。
[Abstract]:The emergence of cloud computing technology has triggered a major change in the field of IT. It has not only changed the mode of delivering software and hardware products by IT, but also affected the way users store and obtain information through the Internet. Cloud computing technology has attracted many service providers to deploy application services to the cloud because of its huge advantages. On the business cloud platform, the efficiency of resource allocation will directly affect the performance and cost of service operation, and it is a key problem that must be solved in the process of commercialization of business cloud platform. Therefore, it is of great significance and application value to study how to allocate resources efficiently on business cloud platform. For this reason, we have learned and learned the basic knowledge and key technologies related to cloud computing, and compared several common open source IaaS cloud platforms. According to the characteristics of each platform and the actual requirements of this study, the OpenStack open source project is selected to build a private business cloud platform, and the multimedia sharing service is deployed to the platform. After learning several common allocation algorithms in cloud computing environment, this paper analyzes and models the resource allocation problem on business cloud platform, and proposes a resource allocation model according to the characteristics of business cloud platform, and implements it. The allocation model is divided into two levels: physical machine level and virtual machine level. At the physical machine level, the simple allocation algorithm of OpenStack project is selected to allocate the physical resources in the platform for each virtual instance. At the virtual machine level, the problem is regarded as a packing problem with variable size, and an algorithm is proposed to realize the resource allocation in the unit of service components. The experimental results show that the proposed resource allocation model achieves the expected allocation effect on the business cloud platform. The allocation results can meet the requirements of service performance, but also reduce some additional overhead, so that the utilization of resources can be more effective.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09
本文编号：2489902
[Abstract]:The emergence of cloud computing technology has triggered a major change in the field of IT. It has not only changed the mode of delivering software and hardware products by IT, but also affected the way users store and obtain information through the Internet. Cloud computing technology has attracted many service providers to deploy application services to the cloud because of its huge advantages. On the business cloud platform, the efficiency of resource allocation will directly affect the performance and cost of service operation, and it is a key problem that must be solved in the process of commercialization of business cloud platform. Therefore, it is of great significance and application value to study how to allocate resources efficiently on business cloud platform. For this reason, we have learned and learned the basic knowledge and key technologies related to cloud computing, and compared several common open source IaaS cloud platforms. According to the characteristics of each platform and the actual requirements of this study, the OpenStack open source project is selected to build a private business cloud platform, and the multimedia sharing service is deployed to the platform. After learning several common allocation algorithms in cloud computing environment, this paper analyzes and models the resource allocation problem on business cloud platform, and proposes a resource allocation model according to the characteristics of business cloud platform, and implements it. The allocation model is divided into two levels: physical machine level and virtual machine level. At the physical machine level, the simple allocation algorithm of OpenStack project is selected to allocate the physical resources in the platform for each virtual instance. At the virtual machine level, the problem is regarded as a packing problem with variable size, and an algorithm is proposed to realize the resource allocation in the unit of service components. The experimental results show that the proposed resource allocation model achieves the expected allocation effect on the business cloud platform. The allocation results can meet the requirements of service performance, but also reduce some additional overhead, so that the utilization of resources can be more effective.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09
【参考文献】
中国期刊全文数据库 前2条
1 李强;郝沁汾;肖利民;李舟军;;云计算中虚拟机放置的自适应管理与多目标优化[J];计算机学报;2011年12期
2 曾智斌;许力;;云计算中高能效的虚拟资源分配策略[J];计算机系统应用;2011年12期
，本文编号：2489902
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2489902.html
