Internet measurement and experimentation platforms, such as PlanetLab, DIMES and Looking Glass, have become essential for network studies and distributed systems evaluation. Despite their many benefits and strengths, an increasingly recognized problem with existing platforms is their inability to capture the geographic and network diversity of the wider, commercial Internet. Lack of diversity and poor visibility into the network hamper progress in a number of important research areas, from network troubleshooting to broadband characterization and Internet topology, and complicate our attempts to generalize from test-bed evaluations of networked systems. The issue has served as motivation for several efforts to build new experimentation platforms and expand existing ones. However, capturing the edge of the network remains an elusive goal. I argue that at its root, the problem is one of incentives. Today's measurement and experimentation platforms adopt two basic incentive models for deployment - cooperative or altruistic - people host the platform in order to use it or they do it for the betterment of science. The shared, if sometimes implicit, assumption is that the goals of those hosting the platforms and the experimenters that use it are aligned. As much of the Internet growth occur in residential broadband and mobile networks, this no longer holds. In this talk, I will present Dasu, an experimentation platform for the Internet's edge that explicitly aligns the objectives of the platform with those of the vantage points, building on broadband characterization as an incentive for adoption. Dasu has been publicly available since mid 2010 and has been adopted by over 95,000 users, across 150 countries. I will then illustrate the value of Dasu' unique perspective with some of the ongoing projects and collaboration already taking advantage of it. |
Back to School Seminar Webpage |