[Colloq] Talk, Monday, November 16 - Krishna Gummadi, Max-Planck Institute

[Rachel Kalweit]

The College of Computer and Information Science Colloquium presents:

Speaker: Krishna Gummadi, Max-Planck Institute for Software Systems in Germany 
Date: Monday, November 16, 2009
Time: 11:00am

Title: Glasnost: Enabling End Users to Detect Traffic Differentiation

Abstract: Holding residential ISPs to their contractual or legal obligations of ``unlimited service'' or ``network neutrality'' is a hard problem because ISP traffic management policies are opaque to end users and governmental regulatory agencies. In this talk, I will describe our experiences with building and deploying Glasnost, a system that improves network transparency by enabling ordinary Internet users to detect whether their ISPs are shaping flows belonging to specific applications.  I will focus on three key challenges we faced: (a) how to keep the barrier of usage low to attract large numbers of users, (b) how to infer traffic differentiation accurately; in particular, how to avoid false accusations of differentiation that adversely affect ISPs' reputation and business, (c) how to keep the system up-to-date with the continuously changing differentiation policies of ISPs world-wide. I will describe how Glasnost addresses each of these challenges in detail.  We have deployed Glasnost in the wild for over a year. During this time more than 350,000 users from over 5,800 ISPs worldwide have used Glasnost to test their ISPs, validating many of our design choices.  I will also show how data from individual Glasnost users can be aggregated to provide regulators and monitors with useful ISP-wide views of the deployment of various differentiation policies.

Bio: Krishna Gummadi leads the Networked Systems research group at the Max-Planck Institute for Software Systems in Germany. He received his Ph.D. (2005) and M.S. (2002) degrees in Computer Science and Engineering from the University of Washington, Seattle. He also holds a B.Tech (2000) degree in Computer Science and Engineering from the Indian Institute of Technology, Madras.

Krishna's research interests are in the measurement, analysis, design, and evaluation of complex Internet-scale systems. His current projects focus on (a) making Internet access infrastructures more transparent, (b) enabling efficient and cost-effective bulk content delivery in the Internet, (c) understanding the evolution of online social network structures and the dynamics of information flows over them, (d) leveraging social networks to design better information sharing systems, and (e) building more trustworthy cloud computing infrastructures. His past work on Internet measurements, peer-to-peer systems, and their workloads has received best paper awards at OSDI, SIGCOMM IMW, and MMCN.