CS Talk - Minlan Yu
Title: Programmable Measurement Architecture for Data Centers
Host: Joan Feigenbaum
Coffee/tea available at 10:15
Abstract:
Measurement serves as the basis for network management. Figuring out what is going on in the network is often harder than deciding what to do about it. However, in today’s networks, measurement is an afterthought. Switch vendors often treat measurement as a second-class citizen, devoting most resources to control functions. Network operators have limited control over what (not) to measure, and have to integrate incomplete measurement data from individual devices. Software-defined networking brings new opportunities for redesigning the network measurement stack to bridge the gap between operator’s measurement requirements and device capabilities. In this talk, I will present a new programmable measurement architecture that supports diverse and dynamic measurement queries.
Our design includes two parts: (1) OpenSketch, a software-defined traffic measurement data plane, which strikes a careful balance between generality (supporting a wide variety of measurement queries) and efficiency (enabling high link speed and low cost). (2) DREAM, a dynamic resource allocator in the control plane that accommodates a large number of measurement queries while maintaining the required accuracy, by leveraging the spatial and temporal multiplexing. Our solutions have gained interests from production data center operators and are in technology transfer to Barefoot networks.
Bio:
Minlan Yu is an assistant professor in the computer science department of University of Southern California. She received her B.A. in computer science and mathematics from Peking University in 2006 and her M.A. and Ph.D. in computer science from Princeton University in 2008 and 2011. After that she was a postdoctoral scholar in UC Berkeley for one year. She has actively collaborated with companies such as Google, AT&T, Microsoft, Facebook, and Barefoot. Her research interests include data networking, distributed systems, enterprise and data center networks, network virtualization, and software-defined networking. She received ACM SIGCOMM doctoral dissertation award in 2012, NSF CAREER award in 2015, and Google research awards in 2013 and 2015.