Advanced Operating Systems (818k)

http://lagoon.cs.umd.edu/classes/818fall09

This web page will be updated as the course goes along. Please check it regularly, and don't forget to reload.


Announcements:


NameOffice HoursEmail
Instructor:Pete Keleherby apptkeleher.umd.edu
Class:Monday/Wednesday, 9:30am - 10:45am, CSI 3120


Overview

The objective is an in-depth understanding of the issues in designing and deploying large-scale decentralized peer-to-peer systems. Secondary themes are understanding reliablity and security as it relates to such systems. The class will consist mostly of lectures by the instructor, a few lectures by students, a midterm and a final, and a medium-scale, individual project. The majority of the lectures will be based on recent research in the literature (such as those at OSDI, NSDI, and Infocom).

Schedule

This is all subject to change, especially the ordering and dates. All of this will be relatively well frozen after students choose their papers to present.

MondayWednesday
2009-08-31Intro/Overview

Pete's rough notes.

2009-09-02Overview, continued...

- Time, Clocks, and the Ordering of Events in a Distributed System
- Wikipedia

Pete's rough notes.

2009-09-07Labor Day2009-09-09Byzantine Agreement.

Pete's rough notes.

2009-09-14- Chord: A Scalable Peer-to-peer Lookup Service for Internet Applications,

- OceanStore: An Architecture for Global-Scale Persistent Storage

- a secure distributed file system using chits

A chord talk. Pete's talk, including first half of Oceanstore, is missing.

2009-09-16Cfs and Ivy.

Pete's talk (actually 2nd half of oceanstore, as this is what we covered).

2009-09-21- Managing Update Conflicts in Bayou: a Weakly Connected Replicated Storage System

- Flexible Update Propagation for Weakly Consistent Replication
by K. Petersen, M. J. Spreitzer, D. B. Terry, M. M. Theimer, and A. J. Demers. Proceedings of the 16th ACM Symposium on Operating Systems Principles (SOSP-16), Saint Malo, France, October 5-8, 1997, pages 288-301.

Pete's talk.

2009-09-23Deno:
- Decentralized Replicated-Object Protocols
Pete Keleher. In The 18th Annual ACM SIGACT-SIGOPS Symposium on Principles of Distributed Computing (PODC), April 1999.

- Deno: A Decentralized, Peer-to-Peer Object Replication System for Weakly-Connected Environments
by U. Cetintemel, P. J. Keleher, B. Bhattacharjee, and M. J. Franklin. In IEEE Transactions on Computers.

Pete's talk.

2009-09-28Pastry:
- Pastry: Scalable, distributed object location and routing for large-scale peer-to-peer systems
A. Rowstron and P. Druschel. IFIP/ACM International Conference on Distributed Systems Platforms (Middleware), Heidelberg, Germany, pages 329-350, November, 2001.

- Storage management and caching in PAST, a large-scale, persistent peer-to-peer storage utility
A. Rowstron and P. Druschel, ACM Symposium on Operating Systems Principles (SOSP'01), Banff, Canada, October 2001
(Jim)

Jim's talk.

2009-09-30CAN:
- A Scalable Content-Addressable Network
Sylvia Ratnasamy, Paul Francis, Mark Handley, Richard Karp and Scott Shenker.

- (mainly the pLSI stuff) pSearch: Information Retrieval in Structured Overlays or here.
(Paulo)

Paulo's talk.

2009-10-05Searching and indexing P2P structures:
- On the Feasibility of Peer-to-Peer Web Indexing and Search
Jinyang Li, Boon Thau Loo, Joe Hellerstein, Frans Kaashoek, David R. Karger, Robert Morris

- Efficient Peer-To-Peer Searches Using Result-Caching
Bobby Bhattacharjee, Sudarshan Chawathe, Vijay Gopalakrishnan, Pete Keleher, Bujor Silaghi
(Koyel)

Koyel's talk.

2009-10-07Censorship-resistence:
- Publius: A Robust, Tamper-Evident, Censorship-Resistant Web Publishing System
Marc Waldman, Aviel D. Rubin and Lorrie Faith Cranor, in Proc. 9th USENIX Security Symposium, 2000.

- Achord: A Variant of the Chord Lookup Service for Use in Censorship Resistant Peer-to-Peer
Steven Hazel and Brandon Wiley, IPTPS 2002.

- Freenet: A Distributed Anonymous Information Storage and Retrieval System
Ian Clarke, Oskar Sandberg, Brandon Wiley, and Theodore W. Hong, in Designing Privacy Enhancing Technologies: International Workshop on Design Issues in Anonymity and Unobservability, LNCS 2009, ed.Hannes Federrath. Springer: New York (2001).

- Introducing Tarzan, A Peer-to-Peer Anonymizing Network Layer
Michael J. Freedman, Emil Sit, Josh Cates and Robert Morris, IPTPS 2002.

Vasileios's talk.

2009-10-12

- What Peer-to-Peer Developers Need to Know about Copyright Law
Fred von Lohmann

- Scooped, again
Jonathan Ledlie, Jeff Shneidman, Margo Seltzer, John Huth

Pete's talk.

2009-10-14Mechanical difficulties
2009-10-19Companies:
- Sinfonia: A New Paradigm for Building Scalable Distributed Systems
(best paper SOSP '97)

- Dynamo: Amazon's Highly Available Key-Value Store
(Xiaoyu)

Xiaoyu's talk.

2009-10-21Google:
- MapReduce: Simplified Data Processing on Large Clusters

- The Google File System
(Paulo)

Paulo's talk.

2009-10-26Midterm2009-10-28OS Scalability:
- FAWN: A Fast Array of Wimpy Nodes

- The Multikernel: A new OS architecture for scalable multicore systems

Paul's talk.

2009-11-02Memory Abstractions:
- Implementing Cooperative Prefetching and Caching in a Global Memory System

- Design Evolution of the EROS Single-Level Store
(Xiaoyu)

Xiaoyu's talk.

2009-11-04I/O:
- Better I/O Through Byte-Addressable, Persistent Memory

- Operating Systems Transactions

Paul's talk.

2009-11-09Distributed Shared Memory:
- Memory Coherence in Shared Virtual Memory Systems

- Implementation and Performance of Munin

- TreadMarks: Distributed Shared Memory on Standard Workstations and Operating Systems

Pete's notes.

2009-11-11LBFS/Intermemory:
- A Low-Bandwidth Network File System

- Towards an Archival Intermemory

(Pete just did a breezy hand-wave through these. That doesn't mean you shouldn't know them, at least in a breezy manner, for the final.)

2009-11-16Block-level access
- Block-Level Security for Network-Attached Disks

- Venti: A New Approach to Archival Data Storage

Adam's talk.

2009-11-18Network file systems - Frangipani: A Scalable Distributed File System

- FARSITE: Federated, Available, and Reliable Storage for an Incompletely Trusted Environment

Vasileios's talk.

2009-11-23Byzantine Stuff:
- Practical Byzantine Fault Tolerance
by Miguel Castro and Barbara Liskov. In Proceedings of the Third Symposium on Operating Systems Design and Implementation (OSDI '99), New Orleans, USA, February 1999.
(Padraig)
2009-11-25OS Security:
- Fabric: A Platform for Secure Distributed Computation and Storage

- Heat-ray: Combating Identity Snowball Attacks using Machine Learning, Combinatorial Optimization and Attack Graphs
(Jim)

2009-11-30- The Part-Time Parliament

- Paxos Made Live
(Padraig)

2009-12-02- Samsara: Honor Among Thieves in Peer-to-Peer Storage

- Cooperative Peer Groups in NICE

2009-12-07Presentations2009-12-09Presentations

Please read the statement on academic integrity.


Requirements

Class requirements include participation in class discussions, presentation of at least one research paper, and a large-scale group project, critiques of other student prentations and projects, and a final exam. Grading will be based on the following:

Midterm:25%
Final:25%
Presentation/Participation:25%
Project:25%