摘要
针对数据中心网络流量大小分布不均匀、传输性能需求不相同的特征,提出了面向传统树型数据中心网络结构的软件定义混合路由机制SHR(software-defined hybrid routing)。SHR通过统计计算将数据流分为大流和小流,为满足其不同的传输性能需求,对大流采用自适应路由算法,对小流采用流量无视路由算法。SHR在Open Flow机制的基础上,将部分控制权从控制器下放至交换机,以减轻网络的额外负载。在Fat-Tree网络拓扑结构上建立流量模型进行性能分析与仿真实验,结果表明,与传统的等价多路径转发ECMP算法相比,SHR能够提高网络吞吐量,降低数据流丢弃率和分组端到端时延,同时减轻网络的额外负载。
In the current data center networks, the flow size distribution is not uniform and the transmission performance requirements of elephant flows and mice flows are different. To address this issue, a software-defined hybrid routing（SHR） scheme was proposed. SHR differentiate data flows by statistical calculations. The elephant flows utilize the adaptive routing algorithm while the mice flows use the oblivious routing algorithm. SHR extends the Open Flow scheme by offloading some basic functions such as flow statistical detection and mice flow forwarding to switches to reduce the switch-controller interaction overhead. Performance evaluations of SHR were carried out using the fat-tree network topology. Results show that SHR can effectively increase network throughput and reduce the flow dropping rate as well as packet delay compared with the traditional ECMP algorithm.
出处
《通信学报》
EI
CSCD
北大核心
2016年第4期44-52,共9页
Journal on Communications
基金
国家自然科学基金资助项目(No.61301119)
教育部高等学校博士学科点专项科研基金资助项目(No.20120191120025)
教育部留学归国人员启动基金资助项目(No.1020607820140002)~~
关键词
云计算
数据中心网络
软件定义网络
路由算法
开放流协议
cloud computing
data center network
software defined network
routing algorithm
OpenFlow
