A practical benchmarking methodology for IPv4 over IPv6 transition technologies

Georgescu Liviu Marius (1251207)

Years from now, IPv4 will only be an important part of the Internet's history, but for now it remains the dominant Internet Protocol and a pending danger to the Internet's expansion. The transition from IPv4 to IPv6 is an ongoing process which will eventually lead to the end of the IPv4 era. This transition has presented the Internet community with numerous challenges, which led to many opportunities for research. Given the variety and complexity of current production networks, a scenario-based approach is one of the major research directions.

The IETF has introduced multiple transition scenarios, among which there is a transition scenario for enterprise networks, where IPv4-capable hosts exist in the IPv6 Internet and network support is required, to convey IPv4 communications through the IPv6 infrastructure. For simplicity, the scenario can be referred as the enterprise 464 scenario. There is a number of suitable transition technologies for this scenario, which can in turn be referred to as 464 technologies. However, a problem remains unsolved. Considering this scenario, which one of these transition technologies is more suitable than the rest? Moreover, different implementations of the same technology can have different capabilities, further complicating the problem.

To support network operators solve this problem, we are proposing a practical benchmarking methodology, exploring various feasibility dimensions of transition technologies. The methodology is associated with a heterogeneous IPv4 and IPv6 network testbed, which we called the IPv6 Network Evaluation Testbed (IPv6NET). To support this methodology, we have used it to analyze the feasibility of two open source transition implementations, covering multiple transition technologies. The feasibility analysis is based on practical means, employing existing implementations of the transition technologies and empirical measurements. To that end we are showing how network performance results, scalability results and operational capability data can be obtained, analyzed and compared. Furthermore, we are presenting a model for building composite indicators, which can identify the optimal transition solution considering certain feasibility dimensions and the specific feasibility needs of enterprise network operators.