標題: SWF: Segmented Wildcard Forwarding for Flow Migration in OpenFlow Datacenter Networks
作者: Kuo, Kuan-Tsen
Wen, Charles H. -P.
Suo, Cheng
Tsai, I-Chen
電機學院
College of Electrical and Computer Engineering
公開日期: 2015
摘要: Networking and performance become major issues in cloud computing. Software Defined Networking (SDN) enables flow -level management and makes routing more effective and flexible for datacenter networks (DCNs). A prior dynamic routing algorithm named Flow Migration shows good results in improving throughput and avoiding congestion through simulation. However, as Flow Migration is applied to a SDN-based DCN, tremendous overhead occurs in modifying flow tables of switches. Therefore, a novel Segmented Wildcard Forwarding (SWF) scheme is proposed to facilitate Flow Migration. SWF leverages the symmetry of a Portland topology and applies wildcard matching for balancing network loading and lowering table -modification cost for flows. Experimental results show that on a Portland 4-8-8 datacenter network, the proposed SWF scheme achieves higher throughput over than the traditional SPB does, and even reduces 70% modification cost used by the original implementation of Flow Migration. As a result, Segmented Wildcard Forwarding (SWF) is proven a mechanism that practically realizes dynamic routing of SDN but also improves network performance as well as cost for flow migration, simultaneously.
URI: http://hdl.handle.net/11536/134357
ISBN: 978-1-4673-6432-4
ISSN: 1550-3607
期刊: 2015 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS (ICC)
起始頁: 313
結束頁: 318
顯示於類別:會議論文