Publications

LiteVisor: A Network Hypervisor to Support Flow Aggregation and Seamless Network Reconfiguration for VM Migration in Virtualized Software-Defined Networks

IEEE Access, JCR 2020 IF 3.367 Top 34.249%, vol. 6, pp. 65945-65959, 2018, doi: 10.1109/ACCESS.2018.2877416.

Gyeongsik Yang, Bong-Yeol Yu, Seong-Mun Kim, Chuck Yoo

Download PDF

Abstract

Network virtualization based on software-defined networking (SDN) has become a necessary technology to provide various services in cloud datacenters. Although many network hypervisors have been proposed to support SDN-based network virtualization, their forwarding techniques excessively consume the limited ternary contents addressable memory of OpenFlow-enabled switches. Moreover, they do not consider network reconfiguration after virtual machine migration. In this paper, we propose LiteVisor, that resolves the two problems mentioned above. It develops the Locator, Identifier, and Tenant sEparating (LITE) scheme. The LITE-based forwarding enables flow aggregation that reduces switch memory consumption. In addition, LiteVisor provides seamless network reconfiguration that does not require tenant controllers to be aware of virtual machine migration based on LITE. We evaluate LiteVisor in terms of the number of flow table entries and network reconfiguration time and compare it with OpenVirteX that is an open-source network hypervisor. The results show that the number of flow rules decreases by up to eight times compared with OpenVirteX in a fat-tree topology. We also demonstrate the seamless network reconfiguration of LiteVisor in the experiments and present the results of the reconfiguration time by the topology size and packet sending interval of hosts.

Paper summary

Separate location from identity when forwarding

A VM can move without requiring every forwarding decision to treat it as a new endpoint.

LiteVisor addresses switch-table growth and network reconfiguration after VM migration. Its LITE forwarding scheme separates locator, endpoint identifier and tenant information.

Keep endpoint identity stable when location changes

LITE forwarding separates the location used to route traffic from the endpoint identity and tenant information used to deliver it. The hypervisor translates virtual flows into this physical representation, allowing intermediate switches to aggregate traffic following the same route. Edge switches still perform encapsulation and decapsulation. When notified of a VM move, LiteVisor recalculates the forwarding path for the new location while hiding that physical change from the tenant controller.

The three forwarding roles separated by LITE.
RolePurpose
LocationRoute through the physical network
Endpoint identityIdentify the destination
Tenant informationMaintain the virtual-network context

Check aggregation and migration as separate outcomes

The OpenVirteX-based prototype uses OpenFlow 1.3 and MPLS. Mininet tests cover a linear topology and three fat-tree sizes. In the linear tests, flow-entry reductions range from about 17% to 75%, depending on host fan-out and intermediate-switch count. Edge rules remain necessary for encapsulation, so aggregation benefits differ by switch layer. Migration tests separately measure network reconfiguration time: it grows with the number of intermediate switches, while the tested packet-sending intervals have relatively little effect.

Flow-entry reduction as intermediate-switch count increases, for three host fan-outs.View full-size figure
In the linear topology, flow-entry savings grow with host fan-out and the number of intermediate switches.

Key takeaways

  1. Separating identity from location benefits both scale and mobility. LITE forwarding supports shared intermediate rules while allowing a moved endpoint’s path to change.
  2. The scope of the benefit is the forwarding network. Entry reductions and migration-path reconfiguration measure two consequences of that separation.