NeTS: Small: Liquid Networking

University Of Colorado At Boulder, Boulder CO

Investigators

Abstract

One of the major advances in information technology infrastructures in the past decade is the use of server virtualization technology. By decoupling the server software from the underlying physical hardware, virtualization helped open up new business models and bring about more efficient computing infrastructures. With virtualization, (virtual) servers are freely able to be started and stopped on demand, and, through live VM migration, even change physical servers at run time without affecting the operating system or applications running on those servers. This dynamic repositioning of servers is used to consolidate servers to save energy, balance load, perform planned maintenance, and optimize user performance (among other benefits). The network, unfortunately, is still tied to the physical resources. This research project introduces liquid networking, which provides the ability to morph a network dynamically, much like migration provided dynamic morphing in the server space. Supporting this raises many challenging research questions that this research project will address. First, understanding how to provide primitives that efficiently transform the network without affecting the correct operation of the network is an important research challenge. Through implementation of multiple applications, a deep understanding of the limitations and capabilities of liquid networking will be explored. With it, new algorithms and tools will need to be developed. Second, understanding the impact future network technology has on liquid networking, namely software-defined networking, and the impact liquid networking will have on future networking technology is an equally important challenge that will be addressed. Broader Impacts: This project will open up an entirely new set of possibilities in the management of a computing infrastructure. With a liquid network, networks will have greater flexibility, lower IT costs, and better security. Additional important broader impacts for society resulting from this project are expected to include enhancing the curriculum of advanced graduate systems courses and enabling undergraduate students, underrepresented minorities and women to participate in the project through programs such as the Discovery Learning Apprentice (DLA) program run by the college of engineering at the University of Colorado, and the Colorado Diversity Initiative.