An Evaluation of Availability, Reliability and Power Consumption for a SDN Infrastructure Using Stochastic Reward Net

Networking infrastructure of a software defined network (SDN) is demanding further studies to achieve continuity and high availability of data transactions for cloud computing services. Nevertheless, the high-speed and complicated network of hosts and network devices often encounters with a variety of failures either of links or system components. This paper aims to study the specific characteristics of and the impact of various failures on a typical SDN infrastructure. We propose a stochastic model using stochastic reward nets (SRN) with the incorporation of hardware failures (of hosts, switches, storages and links) and software failures (virtual machines (VM). The system model is analyzed based on steady state availability in the case of default parameters. Comprehensive sensitivity analyses are conducted to study the system behaviors with regard to different major impacting factors. Reliability analysis is also carried out to pinpoint the role of VM migration in extending the system lifetime. System power consumption based on availability of every module is also conducted to examine the power allocation on system devices and operations. This study provides a helpful basis for network design and implementation in SDN infrastructure.