Daniel Nurmi

Graduate Student, Computer Science University of California Santa Barbara Research Advisor: Rich Wolski VGrADS Activities: Working on Execution System (Fault Tolerance) nurmi@cs.ucsb.edu http://www.cs.ucsb.edu/~nurmi

---

Education

• University of Minnesota
  Studied undergraduate computer science
• Tulane University
  Studied undergraduate computer science in the School of Engineering
• University of Chicago
  MS in computer science, June 2002
• University of California Santa Barbara
  Enrolled in PhD program, computer science

Research Experience

Research Assistant, Fall 2003 - present
Exploring statistical models for use in predicting UNIX
machine failures. Observations currently exploring will
lead to several systems projects (schedulers, optimal
checkpoint interval selection tools, resource managers) that
exploit heavy tailed distributions to make accurate machine
failure predictions in wide area distributed system
environments. Added fault tolerant code and new resource
sensors (file system performance and machine availability
sensors) to successful Network Weather Service software
system. Contributed to and core architect of Augmented
Reality Desktop Window Manager system (ARWin) research
project.

High Performance System Engineer, Winter 1997 - 2003
Responsible for the design and implementation of large and
small scale Linux clusters at ANL. Point of contact for
research, production, and development cluster projects.
Created suite of scalable cluster management software that is
used across wide variety of clusters (Chiba City, Datagrid,
MicroMurals). Worked on several cluster system research
projects involving parallel program execution, scheduling and
resource management (OpenPBS, WIKI), machine configuration and
synchronization (Sanity, Cfg, SciDAC software), and parallel
file systems (PVFS). Active in the community and published
several papers in the field.

Publications

1. J. Brevik, D. Nurmi, R. Wolski, Automatic Methods for Predicting Machine Availability in Desktop Grid and Peer-to-peer Systems, Presentation at CCGrid 2004 GP2PC Workshop, 2004, Chicago, IL
2. D. Nurmi, B. Toonen, Chapter 7 Beowulf Cluster Computing with Linux, 2nd Ed., MIT Press, December, 2004
3. S. Diverdi, D. Nurmi, T. Hollerer, A Framework for Generic Inter-Application Interaction for 3D AR Environments, Presented at IEEE International Augmented Reality Toolkit Workshop, 2003, Tokyo, JP
4. R. Ross, D. Nurmi, A. Cheng, M. Zingale, A Case Study in Application I/O on Linux Clusters, Proceedings of Supercomputing 2001, Denver, CO
5. A. Iamnitchi, D. Nurmi, I. Foster, A Peer to Peer Approach to Resource Location in Grid Environments, Proceedings of the 11th Symposium on High Performance Distributed Computing (HPDC-11), Aug 2002, Edinburgh, UK
6. R. Evard, N. Desai, J. Navarro, D. Nurmi, Clusters as Large-Scale Development Facilities, Proceedings of Cluster 2002, Chicago, IL
7. J. Navarro, N. Desai, R. Evard, D. Nurmi, Scalable Cluster Administration, Proceedings of Cluster 2002, Chicago, IL
8. J. Navarro, R. Evard, D. Nurmi, N. Desai, The Chiba City 1 Scalable Management Architecture, (Article) Newsletter of the IEEE Task Force on Cluster Computing 2002