Abstract: Today, large quantities of data are generated from disparate sources such as users, servers, devices, and sensors dispersed around the globe. The need to store, combine and analyze this information in a meaningful and timely manner has resulted in the need for efficient geo-distributed analytics, creating tradeoffs in terms of cost, timeliness, and quality of results. In this talk, I will discuss the challenges of computing with geo-distributed data, and present some of our work on optimizing computation for such highly-distributed environments. I will present new scheduling algorithms we have developed to optimize aggregating data streams in a geo-distributed analytics setting. In addition, I will present Nebula: a dispersed cloud infrastructure we are building at UMN, that uses voluntary edge resources to enable geo-distributed data-intensive computing.
Abhishek Chandra is an Associate Professor in the Department of Computer Science and Engineering at the University of Minnesota. His research interests are in the areas of Operating Systems and Distributed Systems, with current focus on performance and resource management in Cloud computing, Data-intensive computing, and Mobile computing platforms. He received his B.Tech. in Computer Science and Engineering from IIT Kanpur, India, and M.S. and PhD in Computer Science from the University of Massachusetts Amherst. He is a recipient of the NSF CAREER Award and IBM Faculty Award, his PhD dissertation was nominated for the ACM Dissertation Award, and he is a co-author on multiple Best Paper/Poster Awards.