Michael Golightly will present his research seminar/general exam on Friday May 15 at 10AM in Room 302 (note room). The members of his committee are: Vivek Pai, advisor, Larry Peterson, and Jen Rexford. Everyone is invited to attend his talk, and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. ----------------------------------------------------------------------- Abstract: The PlanetLab global network testbed has grown to consist of over 1000 globally distributed nodes that run hundreds of research experiments concurrently, placing demands on its infrastructure that single- experiment testbeds do not face. While it is widely used in the networking community, very little research has examined the behavior of PlanetLab itself, particularly in the areas of workload and capacity analysis. In light of the number of papers and proposals devoted to resource allocation strategies, auction mechanisms, and testbed fairness, this lack of data is particularly surprising. In order to support continued successful growth of PlanetLab, and to help in planning future testbeds such as GENI, a detailed understanding of how the system is being used is essential. Toward this goal, our work presents a detailed analysis of both the current testbed state as well as some historical analysis using data provided by CoMon, a monitoring service for PlanetLab. We examine node- level behavior as well as experiment-centric behavior to understand how the system and its components are used in practice. We identify root causes of resource contention and workload imbalance, and offer solutions for abatement, finding that targeted investments in infrastructure have the potential to substantially alleviate contention. Through analyzing resource consumption, provision, and disparity, we identify the degree of altruism and selfishness among the users and sites that host PlanetLab nodes. Using this information, we explore the effectiveness of current policies that aim to provide fair sharing of resources. We then explore the appropriateness and feasibility of various resource allocation schemes using a data driven approach based on observed usage patterns. Reading List: [Text Book] [1] L. L. Peterson and B. S. Davie. Computer networks: a systems approach. Morgan Kaufmann, Amsterdam, 4th ed edition, 2007. [Papers] [2] N. Spring, L. Peterson, A. Bavier, and V. Pai. Using planetlab for network research: myths, realities, and best practices. SIGOPS Oper. Syst. Rev., 40(1):17-24, 2006. [3] B. Chun and A. Vahdat. Workload and failure characterization on a large-scale federated testbed. Technical Report IRB-TR-03-040, Intel Research Berkeley, Nov. 2003. [4] K. Park and V. S. Pai. Comon: a mostly-scalable monitoring system for planetlab. SIGOPS Oper. Syst. Rev., 40(1):65-74, 2006. [5] S. Bhatia, A. Kumar, M. Fiuczynski, and L. Peterson. Lightweight, high-resolution monitoring for troubleshooting production systems. In 8th USENIX Symposium on Operating System Design and Implementation (OSDI 2008), December 2008. [6] J. Shneidman, C. Ng, D. C. Parkes, A. AuYoung, A. C. Snoeren, A. Vahdat, and B. Chun. Why markets could (but don't currently) solve resource allocation problems in systems. In HOTOS'05: Proceedings of the 10th conference on Hot Topics in Operating Systems, pages 7-7, Berkeley, CA, USA, 2005. USENIX Association. [7] D. Oppenheimer, J. Albrecht, D. Patterson, and A. Vahdat. Design and implementation tradeoffs for wide-area resource discovery. In 14th IEEE International Symposium on High Performance Distributed Computing, 2005. HPDC-14. Proceedings, pages 113-124, 2005. [8] D. Oppenheimer, B. Chun, D. Patterson, A. C. Snoeren, and A. Vahdat. Service placement in a shared wide-area platform. In ATEC '06: Proceedings of the annual conference on USENIX '06 Annual Technical Conference, pages 26-26, Berkeley, CA, USA, 2006. USENIX Association. [9] Y. Fu, J. Chase, B. Chun, S. Schwab, and A. Vahdat. SHARP: An architecture for secure resource peering. In Proceedings of the nineteenth ACM symposium on Operating systems principles, pages 133- 148. ACM New York, NY, USA, 2003. [10] R. Ricci, D. Oppenheimer, J. Lepreau, and A. Vahdat. Lessons from resource allocators for large-scale multiuser testbeds. SIGOPS Oper. Syst. Rev., 40(1):25-32, 2006.