Analytics on Graphs with a Trillion Edges
Staff - Faculty of Informatics
Start date: 17 March 2015
End date: 18 March 2015
The Faculty of Informatics is pleased to announce a seminar given by Willy Zwaenepoel
DATE: Tuesday, March 17th, 2015
PLACE: USI Lugano Campus, room A23, Red building (Via G. Buffi 13)
TIME: 10.30
ABSTRACT:
Big graphs occur naturally in many applications, most obviously in social networks, but also in many other areas such as biology and forensics. Current approaches to processing large graphs use either supercomputers or very large clusters. In both cases the entire graph must reside in memory before it can be processed. We are pursuing an alternative approach, processing graphs from secondary storage. While this comes with some performance penalty, it makes analytics on very large graphs accessible on a small number of commodity machines. It also has the pleasing property that "if you can store a graph, you can compute on it".
We have developed two systems, one for a single machine and one for a cluster of machines. X-Stream, the single machine solution, aims to make all secondary storage access sequential. It uses two techniques to achieve this goal, edge-centric processing and streaming partitions. X-Stream outperforms the state-of-the-art GraphChi system, because it achieves better sequentiality and because it requires less preprocessing. Slipstream, the cluster solution, starts from the observation that there is little benefit to locality when accessing data from secondary storage over a high-speed network. As a result, partitioning can be dynamic and can focus on achieving load balance, in combination with sequentiality of secondary storage access. The resulting system achieves good scaling behavior and outperforms the state-of-the-art out-of-core Giraph system. With Slipstream we have also been able to process a trillion-edge graph, a new milestone for graph size on a small cluster. I will describe both systems and their performance on a number of benchmarks and in comparison to state-of-the-art alternatives.
BIO:
Willy Zwaenepoel received his BS/MS from the University of Gent, Belgium, and his PhD from Stanford University. He is currently a Professor of Computer Science at EPFL. Before he has held appointments as Professor of Computer Science and Electrical Engineering at Rice University, and as Dean of the School of Computer and Communication Sciences at EPFL. His interests are in operating systems and distributed systems. He is a Fellow of the ACM and the IEEE, he has received the IEEE Kanai Award and several best paper awards, and is a member of the Belgian and European Academies. He has also been involved in a number of startups, including iMimic (now part of Cisco), Midokura and Nutanix.
HOST: Prof. Fernando Pedone
