首页> 外文会议>IEEE Conference on Dependable and Secure Computing >OTMEN: Offloading Traffic Monitoring to Edge Nodes in Software-Defined Datacenter Networks
【24h】

OTMEN: Offloading Traffic Monitoring to Edge Nodes in Software-Defined Datacenter Networks

机译:OTMEN:将流量监控卸载到软件定义的数据中心网络中的边缘节点

获取原文

摘要

Network traffic monitoring is an essential component of many SDN applications, especially in the datacenter environment, where tenants' applications are often distributed over the computing resources. The conventional SDN procedures for flow-level monitoring force the network controller to install fine-grained forwarding rules on switches for the initiated flows, and they rely on the pull-based mechanism to collect traffic statistics from the network switches. Such an approach overwhelms the controller and faces challenges when deployed in datacenter networks, which can have millions of distinct active flows per second. Therefore, in this paper, we introduce the design and implementation of OTMEN, a scalable traffic monitoring system for SDN-enabled datacenter networks. OTMEN decouples the forwarding and monitoring configurations in the data plane to relax the controller, while allowing fine-grained flow-level monitoring at the edge switches. In the OTMEN system, the controller can offload traffic monitoring operations to edge nodes based on the monitoring rules as it provides a rich set of RESTful APIs for monitoring applications. We evaluated our proposed system on a realistic testbed, and the results show that OTMEN provides significant improvements and monitoring overhead reduction compared to the existing solutions.
机译:网络流量监视是许多SDN应用程序的重要组成部分,尤其是在数据中心环境中,租户的应用程序通常分布在计算资源上。用于流级别监视的常规SDN程序迫使网络控制器在交换机上为启动的流安装细粒度的转发规则,并且它们依靠基于拉的机制从网络交换机收集流量统计信息。当部署在数据中心网络中时,这种方法使控制器不堪重负,并面临挑战,后者每秒可能有数百万个不同的活动流。因此,在本文中,我们介绍了OTMEN的设计和实现,OTMEN是用于支持SDN的数据中心网络的可扩展流量监控系统。 OTMEN在数据平面中解耦转发和监视配置,以放松控制器,同时允许在边缘交换机上进行细粒度的流级别监视。在OTMEN系统中,控制器可以基于监视规则将流量监视操作卸载到边缘节点,因为它提供了一组丰富的RESTful API用于监视应用程序。我们在一个实际的测试平台上评估了我们提出的系统,结果表明与现有解决方案相比,OTMEN提供了显着的改进并减少了开销。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号