A Scalable Approach to Alleviating Database Bottlenecks
报告人 : Bruce Maggs is a Professor of Computer Science at Carnegie Mellon UniversityA common three-tiered e-commerce architecture consists of a web server, an application server, and a database. Any of these components may become a performance bottleneck, but cost-effective scalable solutions have been developed only for the web server and (arguably) application server tiers. For example, the web server tier can and is often implemented by a content delivery service. Some operators of these services also offer a platform for running applications, but for several reasons, application services are much more rarely used. In particular, content providers worry about the security of their data and business logic on third-party servers, and may require highly customized run-time environments for their applications. But even if these obstacles are overcome, the database remains a bottleneck, thus reducing the motivation for scaling the application tier. This talk describes S3, a database scalability system based on caching query results that can be applied in several different contexts to augment a database and provide higher throughout for typical e-commerence applications. Among the interesting issues and challenges to be discussed are database consistency, and the tradeoffs between securing data and maintaining a high cache-hit ratio. The talk describes joint work with Natassa Ailamaki, Charrlie Garrod, Phil Gibbons, Amit Manjhi, Todd Mowry, and Anthony Tomasic.
Bruce Maggs is a Professor of Computer Science at Carnegie Mellon University. His research focuses on networks for parallel and distributed computing systems. 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 has been recognized and supported by the National Science Foundation through its National Young Investigator program. He has also held visiting faculty positions at the Massachusetts Institute of Technology and Princeton University. Maggs completed his Ph.D., S.M., and S.B., at M.I.T. under the direction of Charles Leiserson.