|
      
|
|
|
|
In this talk, we describe challenges and solutions to the problem of achieving application reliability in an open distributed system known as the Grid. We discuss an extreme Grid environment composed of voluntary compute nodes under no central authority or control. Such an environment is inherently unreliable in at least two dimensions: correctness and timeliness. In the domain of time, nodes may come and go, fail, and offer unpredictable performance due to compute sharing and network sharing. For applications that require timely execution, this reduces the effectiveness of the Grid. In the domain of correctness, nodes may return incorrect results due to errors, misconfiguration, and malicious behavior, to name a few. The general solution to these problems is replication and consensus. However, approaches to date are ad-hoc with leads either to reduced performance or compromised reliability. We present intelligent replication schemes that can achieve both enhanced reliability and performance. We also address the thorny problem of collusion where nodes may exhibit correlated errors.
Biography and Research Interests
Jon B. Weissman is a leading researcher in the area of Grid computing. His involvement dates back to the influential Legion project at the University of Virginia during his Ph.D. He is currently an Associate Professor of Computer Science at the University of Minnesota where he leads the Distributed Computing Systems Group. His current research interests are in Grid computing, distributed systems, high performance computing, resource management, reliability, and e-science applications. He works primarily at the boundary between applications and systems. He received his B.S. degree from Carnegie-Mellon University in 1984, and his M.S. and Ph.D. degrees from the University of Virginia in 1989 and 1995, respectively, all in computer science. He is a senior member of the IEEE. As well as being a Visitor at the National E-Science Center, Jon has been appointed an Honorary Fellow of the College of Science and Engineering at the University of Edinburgh.
