基于SDN的数据中心网络流量负载均衡的研究与应用
|
摘要
在互联网、云计算弹性扩展,方式灵活的高速推广下,大量的高流量业务快速增长,其中网络中重要体系结构数据中心面临着设备之间负载不均衡,网络带宽使用率底等突出问题.原有网络对流量的控制通常通过网络硬件设备完成,软件定义网络(Software Defined Network,SDN)重要的中心思想是将数据层和网络控制层进行分离,然后通过集中可编程的软件平台控制底层设备,最终灵活实现网络资源按需分布.提出利用SDN优势,选择Ryu作为控制平台,REST API定义相应的模块来对网络数据中心流量的负载均衡进行研究,最后利用仿真模拟平台进行测试,结果表明该策略可以有效地提高网络流量的利用率,减少网络延迟等问题.
With the expansion of the Internet,cloud computing elastic expansion and the rapid promotion of the flexible mode,a large number of high traffic services are growing rapidly.The important architecture data center of the network is facing the outstanding problems of the unbalanced load between devices and the bottom of the network bandwidth.The original network traffic control is usually accomplished through network hardware.The important central idea of Software Defined Network(SDN)is to separate the data layer from the network control layer,and then control the underlying equipment through a centralized and programmable software platform,and finally flexible to realize the distribution of network resources on demand.This paper uses the advantage of SDN,select Ryu as the control platform,REST API defines the corresponding module to study the load balancing of network data center,and finally uses the simulation platform to test it.The result shows that the use of this strategy can effectively improve the utilization of network traffic and reduce network delay.
With the expansion of the Internet,cloud computing elastic expansion and the rapid promotion of the flexible mode,a large number of high traffic services are growing rapidly.The important architecture data center of the network is facing the outstanding problems of the unbalanced load between devices and the bottom of the network bandwidth.The original network traffic control is usually accomplished through network hardware.The important central idea of Software Defined Network(SDN)is to separate the data layer from the network control layer,and then control the underlying equipment through a centralized and programmable software platform,and finally flexible to realize the distribution of network resources on demand.This paper uses the advantage of SDN,select Ryu as the control platform,REST API defines the corresponding module to study the load balancing of network data center,and finally uses the simulation platform to test it.The result shows that the use of this strategy can effectively improve the utilization of network traffic and reduce network delay.
引文
[1]徐涛,丁晓璐,李建伏.K最短路径算法综述[J].计算机工程与设计,2013(11):3900-3905.
[2]陈凡,刘果,李剑锋,等.主要软件定义网络控制器的对比和分析[J].软件,2015(4):97-107.
[3]曾珊,陈刚,齐法制.软件定义网络性能研究[J].计算机科学,2015(6):243-248.
[4]宋海权,郭进,侯孟书,等.基于网络时延的SDN逻辑一致性策略研巧[J].电子科技大学学报,2014(3):730.
[5]左青云,陈鸣,王秀磊,等.一种基于SDN的在线流量异常检测方法[J].西安电子科技大学学报,2015(1):155-160.
[6]吴宇文.基于OpenFlow的网络负载均衡算法的研宄与设计[D].上海:华东师范大学,2014:16.
[7]谢金星,刑文训.网络优化[M].北京:淸华大学出版社,2000:26.
[8]谢蕾.基于SDN的网络流量工程研究[D].成都:电子科技大学,2015:42.
[2]陈凡,刘果,李剑锋,等.主要软件定义网络控制器的对比和分析[J].软件,2015(4):97-107.
[3]曾珊,陈刚,齐法制.软件定义网络性能研究[J].计算机科学,2015(6):243-248.
[4]宋海权,郭进,侯孟书,等.基于网络时延的SDN逻辑一致性策略研巧[J].电子科技大学学报,2014(3):730.
[5]左青云,陈鸣,王秀磊,等.一种基于SDN的在线流量异常检测方法[J].西安电子科技大学学报,2015(1):155-160.
[6]吴宇文.基于OpenFlow的网络负载均衡算法的研宄与设计[D].上海:华东师范大学,2014:16.
[7]谢金星,刑文训.网络优化[M].北京:淸华大学出版社,2000:26.
[8]谢蕾.基于SDN的网络流量工程研究[D].成都:电子科技大学,2015:42.