Towards Self-Managing Large-Scale Distributed Systems

Monday, September 21, 2009

Presenter: Abhishek Chandra
Affiliation: University of Minnesota
Website: http://www.cs.umn.edu/people/faculty/chandra
Time: 11:15 - 12:15
Location: EE/CS 3-125

Abstract

Large-scale distributed systems such as volunteer Grids and clouds consist of large number of shared nodes contributing computational, storage, and network resources for deploying large-scale computational applications. While these systems are attractive due to their scalability and low cost of deployment, they are inherently heterogeneous, unreliable and dynamic, leading to several challenges in their reliable and efficient usage. In this talk, I will present resource management techniques designed towards making these systems self-managing: providing predictable performance and reliability to distributed applications in the presence of such inherent uncertainty. I will focus on techniques developed for two types of distributed platforms: a wide-area distributed Grid environment, and a virtualized cloud computing platform. I will present performance results for some of these techniques obtained through simulations as well as through experiments conducted on a local cluster and PlanetLab: a planetary-scale research testbed.

Bio

Abhishek Chandra is an Assistant Professor in the Department of Computer Science and Engineering at the University of Minnesota-Twin Cities. His research interests are in the areas of Operating Systems, Distributed Systems, and Computer Networks. He received his B.Tech. degree in Computer Science and Engineering from IIT Kanpur in 1997, and M.S. and PhD degrees in Computer Science from the University of Massachusetts Amherst in 2000 and 2005 respectively. His PhD dissertation titled "Resource Allocation for Self-Managing Servers" was nominated for the ACM Dissertation Award in 2005. Prof. Chandra won the NSF CAREER Award in 2007, and was a co-author on the Best Student Paper at IEEE ICAC'05.