Fault Tracking Framework for Software-Defined Networking (SDN)

Introduction

The emergence of software-defined networking (SDN) brings out opportunities to both industry and academia to improve the functionalities in networking and cut down the cost of networking operation too. With this, SDN raises a set of fundamental questions, including architectural issues like whether control should be centralized or distributed, and whether control and data planes should be separated. “SDN starts from two simple ideas: generalize network hardware so it provides a standard collection of packet processing functions instead of a fixed set of narrow features, and decouple the software that controls the network from the devices that implement it. This design makes it possible to evolve the network without having to change the underlying hardware and enables expressing network algorithms in terms of appropriate abstractions for particular applications” (Casado, M.,2014). The SDN concepts are new and still under research, and standard development stage in academic while industries are working to develop product under trial, but not in position to commercialize them. From one perspective, it is visible from different initiatives how SDN is able to foster innovation in networks at a level that has not been seen in the last decade. On the other hand, industries are usually too concerned with time to market rather than providing conceptually elegant solutions (Wickboldt, J. A., 2015). Particularly, the majority of efforts are focused on providing solutions and services over SDN networks, whereas network management specifically fault management is not given much attention. However, as widely recognized, management must not be an afterthought (Schonwalder, J., et al., 2009).

The OpenFlow (ONF) was first introduced in March 2008, from the network research group of Stanford University and collaborators. The first time the concept of “network operating system (NOX)” appeared in this context was through the proposal of the NOX controller (Cisco, 2013). NOX provided a concept how to enable the development of software to control OpenFlow-based networks. The ONF is also responsible for the maintenance of OpenFlow standards, technical specifications of the OpenFlow Switch, and the conformance testing of SDN enabled devices. The architecture of SDN consists of three layers: the application layer, the control layer, and the infrastructure layer. The function of the two lower layers are called OpenFlow controller and OpenFlow Switch, corresponding to the control and data planes of traditional IP/MPLS network switches and routers. With the OpenFlow standard, the OpenFlow controller instructs the OpenFlow Switch to define the standard functional messages such as packet-received, send-packet-out, modify-forwarding-table, get-stats, etc. (Sun, S., 2015).

In addition to the ONF, the Internet Engineering Task Force (IETF), the International Telecommunications Union Telecommunication Standardization Sector (ITU-T), the European Telecommunications Standards Institute (ETSI), and the China Communications Standards Association (CCSA) have also started standardization work on SDN. IETF issued an RFC concerning the requirements and application issues relating to SDN from an operator’s perspective (IETF, 2014), while ITU-T has not published a formal recommendation since the project began in 2012. Table 1 summarizes the formally published SDN standards (Sun, S., 2015).

Sun, S., 2015.