一种云环境下集群弹性负载均衡机制的研究

发布时间：2018-06-04 09:25

本文选题：云计算 + 虚拟机　；参考：《太原理工大学》2013年硕士论文

【摘要】：随着现代计算机硬件技术的发展,使得大规模计算成为可能,在此基础上,近些年来提出了许多新的计算模式,云计算便是其中之一。有了云计算后用户以廉价的成本换取需要的IT基础设施服务,例如计算资源、存储资源等。云计算以方便、快捷、廉价的租用方式对IT领域产生了重大的影响。弹性云计算利用虚拟化技术,将一组分布的硬件资源集中管理后对外提供服务,其中需要解决的一个重要问题是负载均衡问题,负载均衡机制的优劣与云计算平台所提供的服务质量有着密切的联系。在传统的集群环境下,任务处理单元是硬件物理主机,如果主机添置过多,不仅会导致运行维护成本过高,而且当负载过低时,计算资源利用率过低使得造成浪费；如果服务节点配置过少,负载超过总承载能力,会对QoS造成极大的影响,所以传统的负载均衡是一种刚性负载均衡。而云环境中的虚拟机资源是一种弹性计算资源,可以根据需求动态的按需调节,当需要添加一台虚拟机时,向云控制中心请求分配一台满足配置需求的虚拟机,当虚拟机使用完毕时,云控制中心将会回收。在以往的研究中忽略了这一特性。针对上述问题,本文提出了在云环境下,针对不同负载能力的需求,通过个性化的计算资源配置,结合负载预测的方法,创建具有一定数量的、不同类型配置虚拟机的资源池弹性负载均衡机制。在该机制下,当提供服务的虚拟集群负载过高时,向资源池管理器请求一台满足配置的虚拟机加入服务集群；当提供服务的虚拟集群负载过低时,就选择一台虚拟机离线以降低系统资源的消耗。同时本文借鉴“绿色计算”的思想,降低云计算中心的电能消耗,提高整个计算资源的利用率,进而就减少了向自然环境中排放的温室气体,达到了环保的绿色的目的。
[Abstract]:With the development of modern computer hardware technology, large-scale computing is possible. On the basis of this, many new computing models have been proposed in recent years, among which cloud computing is one of them. With cloud computing, users can exchange cheap costs for the required IT infrastructure services, such as computing resources, storage resources, and so on. Cloud computing has had a significant impact on IT in a convenient, fast and cheap way to rent. Flexible cloud computing uses virtualization technology to centralize a group of distributed hardware resources to provide services to the outside world. One of the important problems that need to be solved is load balancing. Load balancing mechanism is closely related to the quality of service provided by cloud computing platform. In the traditional cluster environment, the task processing unit is the hardware physical host, if the host is added too much, it will not only lead to the running and maintenance costs too high, but also when the load is too low, the low utilization of computing resources will lead to waste. If the service node configuration is too small and the load exceeds the total load capacity, it will have a great impact on QoS, so the traditional load balancing is a kind of rigid load balancing. The virtual machine resource in the cloud environment is a kind of flexible computing resource, which can be dynamically adjusted according to the demand. When a virtual machine needs to be added, a virtual machine can be allocated to the cloud control center to meet the configuration requirements. When the virtual machine is finished, the cloud control center will be recycled. This feature has been ignored in previous studies. In order to solve the above problems, this paper puts forward that in the cloud environment, according to the demand of different load capacity, through the personalized computing resource allocation, combined with the method of load prediction, we create a certain number of, Resource pool elastic load balancing mechanism for different types of virtual machines. Under this mechanism, when the load of the virtual cluster providing the service is too high, a virtual machine that meets the configuration is requested to join the service cluster by the resource pool manager; when the load of the virtual cluster providing the service is too low, Select a virtual machine offline to reduce system resource consumption. At the same time, this paper uses the idea of "green computing" to reduce the power consumption of cloud computing center, improve the utilization rate of the whole computing resources, and then reduce the greenhouse gas emissions to the natural environment, and achieve the goal of environmental protection green.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP302

【相似文献】