Tenant Orchestration in a Neutral Cloud

November 10, 2015
10:30am
CS Conference Room
Speaker: George Kesidis, Professor, School of EECS, Penn State

Abstract

Customer uptake of cloud services is continuing to increase, so that one can envision a near-future public cloud that is much more efficient, profit minded, and neutral, while its tenants will be increasingly price sensitive and committed to cloud services. We describe a general framework for the management by a large, long-lived tenant of its cloud resources subjected to neutral SLAs, i.e., limitations on information and cooperation between cloud and tenant. The tenant may exploit a hierarchy of control knobs to optimize a hierarchy of benefit-cost criteria based on a near-term forecasts of effective resource capacities or prices and of its own workload demand. Some of our recent results in this general area will be discussed, including  an analysis of multi-resource aggregative games among competing tenants, and how a cloud might employ simple chance constraints to consolidate workload and exploit statistical multiplexing.  We conclude with an overview of our recent research on energy-efficient cloud operations. This work is in collaboration with Bhuvan Urgaonkar, IBM Research, and students.

Speaker Bio

George Kesidis received his MS (in 1990) and PhD (in 1992) from UC Berkeley in EECS. Following eight years as a professor of ECE at the University of Waterloo, he has been a professor of CSE and EE at the Pennsylvania State University since 2000. His research interests include many aspects of networking, cyber security and machine learning, particularly intrusion detection based on large-scale network datasets, and, more recently, energy efficiency and the impact of economic policy. His work has been supported by over a dozen NSF research grants and several Cisco Systems URP gifts, the latter supporting applied work on cyber security.  His web site is http://www.cse.psu.edu/~gik2


500 W. 120th St., Mudd 1310, New York, NY 10027    212-854-3105               
©2014 Columbia University