Experimenting with a Content Delivery Network

Bruce Maggs, Carnegie Mellon University

This talk describes the use of a content delivery network as a testbed for experimental networking. In particular, it presents two recent studies that were performed using Akamai’s network of servers. The network is unique in two ways. First, the scale and diversity of server deployment is unmatched: over 14,000 servers on over 2000 networks in over 70 countries. Second, the network collects voluminous data in the course of providing content delivery services to over 1000 customers. The first study examines the benefits of “multihoming”, i.e., the practice of subscribing to multiple internet service providers. Multihoming was originally employed by stub networks to enhance the reliability of their network connectivity. With the advent of commercial “intelligent route control” products, stubs can now also leverage multihoming to improve performance. This study aims to quantify the extent to which multihoming can improve reliability and performance. The second study is an attempt to characterize a significant fraction of all interdomain http traffic flows. The key idea is that for each request received by an Akamai server for an image that appears on a customer’s web page, it is possible to infer a request to the customer’s origin server from the same client for that page. The latter requests, from clients worldwide to servers operated by content providers, constitute a large body of interdomain flows.

Speaker Biography

Bruce Maggs received the S.B., S.M., and Ph.D. degrees in computer science from the Massachusetts Institute of Technology in 1985, 1986, and 1989, respectively. After spending one year as a Postdoctoral Associate at MIT, he worked as a Research Scientist at NEC Research Institute in Princeton from 1990 to 1993. In 1994, he moved to Carnegie Mellon, where he is now an Associate Professor in the Computer Science Department. While on a two-year leave-of-absence from Carnegie Mellon, Maggs helped to launch Akamai Technologies, serving as its Vice President for Research and Development, before returning to Carnegie Mellon. He retains a part-time role at Akamai as Vice President for Research.

Maggs’s research focuses on networks for parallel and distributed computing systems. In 1986, he became the first winner (with Charles Leiserson) of the Daniel L. Slotnick Award for Most Original Paper at the International Conference on Parallel Processing, and in 1994 he received an NSF National Young Investigator Award. He was co-chair of the 1993-1994 DIMACS Special Year on Massively Parallel Computation and has served on the program committees of SPAA, SODA, STOC, PODC, and many other technical conferences.