A Demonstration Scenario for the NorNet Resilient Network Testbed
Due to the availability of reasonably inexpensive Internet connections, cloud computing is becoming more and more popular and widespread. Instead of maintaining servers locally, services are provided by remote virtual servers at data centres. In the recent years, a lot of research work has been done in the area of server redundancy and service migration. However, for many cloud computing users today, the access to the Internet constitutes a single point of failure. That is, if such an Internet connection is broken, all cloud services become unavailable. Therefore, further research on network resilience is necessary. The Simula Research Laboratory has therefore established the NorNet project, in order to create a multi-site research testbed platform for network resilience experiments. Particularly, all NorNet sites are multi-homed, i.e. connected to at least two Internet Service Providers (ISP).
The TCP protocol, the most widespread Transport Layer protocol in the Internet, has no support for multi-homing. Therefore, the Stream Control Transmission Protocol has been developed and standardised by the Internet Engineering Task Force (IETF) as RFC 4960. Furthermore, the IETF has standardised Reliable Server Pooling (RSerPool, see RFC 5351). It constitutes a Session Layer framework for server redundancy. RSerPool is based on the SCTP protocol, i.e. the underlying data transport supports multi-homing. However, due to the lack of a larger-scale, multi-homed test setup, it has so far not been widely evaluated in real-world, Internet-based, multi-homed setups. For the RSerPool reference implementation, only a demonstration tool intended for single-homed systems is available yet.
In order to illustratively demonstrate the resilience functionalities of SCTP and RSerPool in a multi-site, multi-homed Internet setup, the subject of this masteroppgave will be to design a demonstration platform for that purpose. That is, based on the existing RSerPool implementation and the demonstration software, a NorNet-based demonstration scenario will be designed and developed. The demonstration setup is particularly intended to also show the basic functionalities to an audience without a broader knowledge of the networking backgrounds. However, since the demonstration setup will for the first time combine these functionalities in a larger-scale Internet setup, it will also be an evaluation of the underlying protocols and their implementations as well as a proof of concept for the application software. It is also intended that improvements to the underlying Open Source implementations will be published, in order to make them also available to other users.
This project is *your* chance to get involved into state-of-the-art network software development and worldwide IETF protocol standardisation as well as into international, top-level research at the Simula Research Laboratory in Fornebu! Are you interested in this challenge? Do not hesitate to contact us!
For more information please contact Thomas Dreibholz (email@example.com).
What you should know:
- You should have some programming experience, preferably under Linux/Unix.
What you will do:
- Design a useful and illustrative demonstration setup, based on the NorNet resilient network testbed.
- Develop the necessary demonstration software.
- Run some experiments with the demonstration interface.
What you will learn:
- You will learn about resilient networks and applications.
- You will get to know the SCTP protocol, a state of the art transport protocol.
- You will learn about network programming.
- Dreibholz, Thomas et al: ``The RSPLIB Project – From Research to Application´´
- Lei, Peter et al: ``An Overview of Reliable Server Pooling Protocols´´
- Dreibholz, Thomas: ``Reliable Server Pooling – Evaluation, Optimization and Extension of a Novel IETF Architecture´´
- Dreibholz, Thomas: ``Evaluation and Optimisation of Multi-Path Transport using the Stream Control Transmission Protocol´´
- NorNet -- A Programmable Testbed for Measurements and Experimental Networking Research.