CS739 Spring 2006: Reading List Measurement -- An Analysis of Internet Content Delivery Systems : himani Stefan Saroiu, Krishna P. Gummadi, Richard J. Dunn, Steven D. Gribble, and Henry M. Levy, (University of Washington) OSDI'02 Question: Describe the methodology employed in this paper. What traffic could be missed with this approach? Are there any statistics that the authors didn't report that you wish they had gathered? Do you think that access to the complete data would lead to different conclusions than those drawn by the authors? Survey -- Distributed Operating Systems : suresh Andrew S. Tanenbaum and Robbert Van Renesse ACM Computing Surveys, Volume 17, Issue 4 (December 1985) Question: This paper surveys distributed systems as of 1985. What were the goals of these distributed systems? What were the assumptions (in terms of workload and environment) of these systems? What was the state of the art for distributed systems at this time? SpriteMigration -- Transparent Process Migration: Design Alternatives and the Sprite Implementation : meenali Fred Douglis and John K. Ousterhout Software - Practice and Experience, Volume 21, Number 8, 1991, Pages 757-785. Question: The Sprite migration mechanism makes trade-offs between four factors: transparency, residual dependencies, performance, and complexity. What did the Sprite designers choose for each factor? Do you think they made appropriate trade-offs? Why or why not? Zap -- The Design and Implementation of Zap: A System for Migrating Computing Environments : tannenba Steven Osman, Dinesh Subhraveti, Gong Su, and Jason Nieh, Columbia University, OSDI'04 Question: Compare Zap to Sprite migration. You can discuss goals and assumptions, as well as design or implementation features of the systems themselves. VMmigration -- Live Migration of Virtual Machines Christopher Clark, Keir Fraser, and Steven Hand, University of Cambridge Computer Laboratory; Jacob Gorm Hansen and Eric Jul, University of Copenhagen; Christian Limpach, Ian Pratt, and Andrew Warfield, University of Cambridge Symposium on Networked Systems Design and Implementation (NSDI'05), May 2005 Question: Compare migration in Xen to Sprite migration. You can discuss goals and assumptions, as well as design or implementation features of the systems themselves. Time -- Time, Clocks, and the Ordering of Events in a Distributed System Leslie Lamport CACM, July 1978, vol 21, no 7. Snapshots -- Distributed Snapshots: Determining Global States of Distributed Systems K. Mani Chandy and Leslie Lamport ACM Transactions on Computer Systems, vol 3, no. 1, February 1985 Question: Create your own question for Time and Snapshots. SpriteFS -- Caching in the Sprite network file system Michael N. Nelson, Brent B. Welch, John K. Ousterhout ACM Transactions on Computer Systems (TOCS) Volume 6 , Issue 1 (February 1988) Pages: 134 - 154 Question: The client-side caching policy in a distributed file system impacts many factors, such as consistency, performance, and the amount of state the server must maintain. Discuss how the design choices made in Sprite impact these factors. Petal -- Petal: Distributed Virtual Disks Ed Lee, Chandramohan Thekkath ASPLOS 7, October 1996. Question: How is Petal similar to RAID storage? How is Petal different? (Goals, assumptions, interface, and design issues are all fine topics to address.) Frangipani -- Frangipani: A Scalable Distributed File System Chandramohan Thekkath, Tim Mann, Ed Lee SOSP 16, October 1997 Frangipani Question: What do you think is the most interesting (or most relevant or most important) contribution of the research described in this paper? xFS : Serverless Network File Systems : himani Tom Anderson, Mike Dahlin, Jeanna Neefe, David Patterson, Drew Roselli, Randy Wang. SOSP 15, December 1995. Question: How does xFS utilize a log for data and meta-data? You can discuss any set of relevant issues, such as: What is the purpose of the log? What are the advantages of writing to a log? How are the data structures maintained? How does the use of a log in xFS differ from that of Petal and/or Frangipani? TACC -- Cluster-Based Scalable Network Services A. Fox, S. Gribble, Y. Chawathe and E. A. Brewer. Proceedings of SOSP '97, St. Malo, France, October 1997. Question: What do you think is the most interesting (or most relevant or most important) contribution of the research described in this paper? GoogleFS -- The Google File System Sanjay Ghemawat, Howard Gobioff, Shun-Tak Leung SOSP'03 Question: The TACC paper showed some of the benefits of relying upon soft state and/or stale information in the presence of node failures. Where does GoogleFS rely upon soft state and stale information? Discuss the implications and whether or not these appear to be good design decisions. MapReduce -- MapReduce: Simplified Data Processing on Large Clusters Jeffrey Dean and Sanjay Ghemawat OSDI'04 Question: How does the MapReduce programming model interact with the Google file system (GFS)? Where does MapReduce use GFS and where does it not? What are the performance and reliability implications of using GFS or not? Byzantine -- The Byzantine Generals Problem Leslie Lamport, Robert Shostak, and Marshall Pease ACM Transactions on Programming Languages and Systems, Vol 4, No. 3, July 1982 No write-up needed for this paper. LBFS -- A Low-Bandwidth Network File System Athicha Muthitacharoen, Benjie Chen (MIT), David Mazieres (NYU), SOSP'01 Question: What is the basic approach that LBFS uses to improve performance? Describe precisely the types of workloads that should perform well with LBFS. How would you construct a benchmark or workload to stress interesting behavior in LBFS?