基于云计算的负载均衡技术的研究
发布时间：2018-08-28 11:18
【摘要】：负载均衡作为云计算的关键技术之一[1]，可以把云负载动态均衡地分配到云系统所有的节点中。有效地负载均衡算法在提高用户满意度、提高资源利用率和避免虚拟机频繁迁移起着重要的作用。 本文对基于云计算的负载均衡技术进行了研究，在云资源提供策略、云计算链式负载模型及监控机制和云任务调度等方面提出了相应的策略，主要工作包括： 1.分析了虚拟机部署机制、引起虚拟机性能下降因素和云环境链式负载模型，提出了差异化资源池分配策略，具体包括物理资源池差异搭建策略和虚拟机资源池差异部署策略。仿真实验结果表明，与现有的资源池搭配策略相比，该策略能够充分利用云资源、提高资源利用率。 2.基于现有的云计算环境中的负载均衡技术，，量化了云环境中多种资源的使用情况，以细资源粒度方式定义了云计算中不同的负载均衡类型，提出了一种可反馈的负载均衡策略及双监控器策略。通过仿真实验证明其能有效的监控资源使用情况，使云系统维持相对负载均衡状态。 3.对比分析了现有的任务调度策略，说明了任务类型分类方法和任务资源需求量计算方法，提出了基于负载均衡的任务调度算法，并从响应时间的角度对该算法进行了改进。与现有的基于负载均衡的任务调度算法对比，实验仿真证明了该算法在系统的整体负载均衡方面和响应时间方面具有一定优势。
[Abstract]:As one of the key technologies of cloud computing, load balancing can be dynamically distributed to all nodes in cloud system. Effective load balancing algorithm plays an important role in improving user satisfaction, improving resource utilization and avoiding frequent migration of virtual machines. In this paper, the load balancing technology based on cloud computing is studied, and the corresponding strategies are put forward in the aspects of cloud resource supply strategy, cloud computing chain load model, monitoring mechanism and cloud task scheduling. The main work includes: 1. This paper analyzes the mechanism of virtual machine deployment, causes the degradation of virtual machine performance and the chain load model of cloud environment, and puts forward a strategy for allocation of differentiated resource pool, including physical resource pool differential construction strategy and virtual machine resource pool differential deployment strategy. The simulation results show that compared with the existing resource pool collocation strategy, this strategy can make full use of cloud resources and improve resource utilization. 2. Based on the existing load balancing technology in cloud computing environment, this paper quantifies the use of various resources in cloud environment, and defines different load balancing types in cloud computing in a fine-resource granularity manner. A feedback load balancing strategy and a dual monitor strategy are proposed. The simulation results show that it can effectively monitor the use of resources and make the cloud system maintain relative load balance. This paper compares and analyzes the existing task scheduling strategies, explains the task classification method and the task resource requirement calculation method, proposes a task scheduling algorithm based on load balancing, and improves the algorithm from the point of view of response time. Compared with the existing task scheduling algorithms based on load balancing, the simulation results show that the algorithm has some advantages in the overall load balancing and response time of the system.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09
[Abstract]:As one of the key technologies of cloud computing, load balancing can be dynamically distributed to all nodes in cloud system. Effective load balancing algorithm plays an important role in improving user satisfaction, improving resource utilization and avoiding frequent migration of virtual machines. In this paper, the load balancing technology based on cloud computing is studied, and the corresponding strategies are put forward in the aspects of cloud resource supply strategy, cloud computing chain load model, monitoring mechanism and cloud task scheduling. The main work includes: 1. This paper analyzes the mechanism of virtual machine deployment, causes the degradation of virtual machine performance and the chain load model of cloud environment, and puts forward a strategy for allocation of differentiated resource pool, including physical resource pool differential construction strategy and virtual machine resource pool differential deployment strategy. The simulation results show that compared with the existing resource pool collocation strategy, this strategy can make full use of cloud resources and improve resource utilization. 2. Based on the existing load balancing technology in cloud computing environment, this paper quantifies the use of various resources in cloud environment, and defines different load balancing types in cloud computing in a fine-resource granularity manner. A feedback load balancing strategy and a dual monitor strategy are proposed. The simulation results show that it can effectively monitor the use of resources and make the cloud system maintain relative load balance. This paper compares and analyzes the existing task scheduling strategies, explains the task classification method and the task resource requirement calculation method, proposes a task scheduling algorithm based on load balancing, and improves the algorithm from the point of view of response time. Compared with the existing task scheduling algorithms based on load balancing, the simulation results show that the algorithm has some advantages in the overall load balancing and response time of the system.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09
【参考文献】
相关期刊论文 前10条
1 刘之家;;一种基于云计算的负载均衡技术的研究[J];广西师范学院学报(自然科学版);2011年02期
2 李陶深;张希翔;;云计算下区分服务的演化博弈调度算法[J];北京邮电大学学报;2013年01期
3 程朝辉,李伟,涂亚明;软件静态度量和预测模型研究[J];计算机工程与应用;2005年24期
4 李建锋;彭舰;;云计算环境下基于改进遗传算法的任务调度算法[J];计算机应用;2011年01期
5 张建勋;古志民;郑超;;云计算研究进展综述[J];计算机应用研究;2010年02期
6 郑洪源;周良;吴家祺;;WEB服务器集群系统中负载平衡的设计与实现[J];南京航空航天大学学报;2006年03期
7 陈廷伟;周山杰;秦明达;;面向云计算的任务分类方法[J];计算机应用;2012年10期
8 左利云;曹志波;;云计算中调度问题研究综述[J];计算机应用研究;2012年11期
9 杜W
本文编号：2209238
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2209238.html
