首页> 外文会议>International Computer Conference, Computer Society of Iran >Robust Software Defined Networks Through Dynamic MiddleBox Selection
【24h】

Robust Software Defined Networks Through Dynamic MiddleBox Selection

机译:强大的软件通过动态中间盒选择定义了网络

获取原文

摘要

In the current networks, multiple servers and Middleboxes (MBs) are applied to enhance the network performance and availability. Therefore, different load-balancing strategies are in use to distribute network traffic among these devices. For resource utilization optimization, the status of all network devices should be considered in the traffic distribution procedure. Software-defined Networking (SDN) has an important role in maximizing network resource utilization, availability, and robustness. In this paper, we implemented a "resource-based" load-balancing method in SDN that considers the MBs' status during the dynamic routing to avoid overloaded MBs in SDNs. Overloaded devices can cause network unavailability and vulnerability. To observe the efficiency of the proposed solutions, we used mininet emulator, and we extracted the CPU, RAM, and bandwidth usage of each MB during the test time. The obtained results show that our proposed method improves the availability and resource utilization of the tested network approximately 85% and 95% in comparison to Round-Robin (RR) and IP-Hash based approaches, respectively. The results also declared that bandwidth and CPU are the most crucial resources of the tested MBs.
机译:在当前网络中,应用多个服务器和中间盒(MBS)来增强网络性能和可用性。因此,不同的负载平衡策略用于在这些设备之间分配网络流量。对于资源利用优化,应在流量分发过程中考虑所有网络设备的状态。软件定义的网络(SDN)在最大化网络资源利用率,可用性和稳健性方面具有重要作用。在本文中,我们在SDN中实现了“基于资源的”负载平衡方法,该方法在动态路由期间考虑MBS的状态,以避免SDN中的超载MB。超载设备可能导致网络不可用和漏洞。要遵守所提出的解决方案的效率,我们使用了MinInet仿真器,并在测试时间内提取了每个MB的CPU,RAM和带宽使用。所得结果表明,与循环(RR)和基于IP-Hash的方法相比,我们所提出的方法可以改善测试网络的可用性和资源利用率约为85%和95%。结果还宣布带宽和CPU是测试MBS最重要的资源。

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