Non-tunneling Overlay Approach for Virtual Tenant Networks in Cloud Datacenter

Ryota KAWASHIMA  Hiroshi MATSUO  

Publication
IEICE TRANSACTIONS on Communications   Vol.E97-B   No.11   pp.2259-2268
Publication Date: 2014/11/01
Online ISSN: 1745-1345
DOI: 10.1587/transcom.E97.B.2259
Type of Manuscript: Special Section PAPER (Special Section on Network Virtualization, and Fusion Platform of Computing and Networking)
Category: 
Keyword: 
SDN,  OpenFlow,  network virtualization,  overlay,  datacenter,  

Full Text: PDF(2MB)>>
Buy this Article




Summary: 
Network virtualization is an essential technology for cloud datacenters that provide multi-tenancy services. SDN-enabled datacenters have introduced an edge-overlay (distributed tunneling) model to construct virtual tenant networks. The edge-overlay model generally uses L2-in-L3 tunneling protocols like VXLAN. However, the tunneling-based edge-overlay model has some performance and compatibility problems. We have proposed a yet another overlay approach without using IP tunneling. Our model leverages two methods, OpenFlow-based Virtual/Physical MAC address translation and host-based VLAN ID usage. The former method replaces VMs' MAC addresses to physical servers' ones, which prevents frame encapsulation as well as unnecessary MAC address learning by physical switches. The later method breaks a limitation of the number of VLAN-based virtual tenant networks (4094) by allocating entire VLAN ID space to each physical server and by mapping VLAN ID to VM with OpenFlow controller support. In our model, any special hardware equipment like OpenFlow hardware switches is not required and only software-based virtual switches and the controller are used. In this paper, we evaluated the performance of the proposed model comparing with the tunneling model using 40GbE environment. The results show that the performance of VM-to-VM communication with the proposed model is close to that of physical communication and exceeds 10Gbps throughput with large TCP segment, and the proposed model shows better scalability for the number of VMs.