CPMan: Adaptive control plane management for software-defined networks

Abstract

Software-Defined Network (SDN) is an emerging network paradigm which enables flexible network management by separating control plane from data plane. Due to the centralized management scheme that SDN adopts, intensive control plane overhead incurs as the scale of SDN increases. One way of alleviating the overhead is to either distribute the overhead to multiple controllers, or offload the overhead to switches. However, both of the approaches require modification on SDN de-facto standard, therefore, not viable for practical use. As an alternative solution, we propose a new Control Plane Management (CPMan) method without changing the underlying SDN protocol. The key idea of the proposed method is to maintain as much information as possible inside the switches, so that with enriched information, the switches may less frequently query the controller. However, it comes with an over-utilization of switch memory as tradeoff; therefore, deserves a careful management. Unfortunately, it is non-trivial problem to control the switch memory utilization under the designated threshold (e.g., switch capacity), due to the absence of a detailed correlation model between memory utilization and the affecting parameters. To resolve this issue, we adopt a lightweight feedback loop-based control scheme to adaptively tune the affecting parameters to minimize the control plane overhead, while avoiding the switch memory exhaustion. We design and implement the proposed method as an SDN application and evaluate its performance in an emulated network.