CS262B Spring 2006

Eric Brewer and Timothy Roscoe

ScheduleJanuary 17 -- May 9, Tu/Th 12:30-2 in 306 Soda (CCN 26797)
Prerequisites 262A (or permission)
Text None
Handouts Papers will be handed out or online
Grades 85% group project, no midterm, 15% reading summaries and problem sets
Projects Groups of 2-3 will complete small research projects that involve some implementation and evaluation. Groups will be (primarily) self selected and have significant control over their project. We have several projects to suggest as well.
Reading summariesOne-half page at most with at least one criticism due by beginning of class to cs262profs AT db.cs.berkeley.edu.
Contact Info Prof. Brewer : 623 Soda Hall, (510) 642-8143
     Office hours: TBA
Timothy Roscoe: Intel Research Berkeley, 2150 Shattuck Ave, Suite 1300, (510) 495-3086
     Office Hours: Tu 11-12 & Wed 11-12 at Intel, or by appointment


Poster Session: May 11, 2006, 1:30-3:30, in the 6th floor atrium (Soda)
Papers due May 19th at 4pm via e-mail


Date Tuesday Date Thursday
1/17 Statistics

Excel spreadsheet, slides
1/19 Exterminate All Operating System Abstractions, Engler and Kaashoek
Extensible Kernels are Leading OS Research Astray, Druschel et al.
1/24 CAP Theorem [no summary]
The Case for NOW
1/26 Cluster Hash Table
Cluster-Based Scalable Network Services
1/31 Google File System
MapReduce: Simplified Data Processing on Large Clusters"
2/2 Lessons from Giant-Scale Services
2/7 Coda 2/9 Flexible Update Propagation for Weakly Consistent Replication
Managing Update Conflicts in Bayou
[Summary Required only for the latter paper]
2/14 PRACTI Replication [new version] 2/16 Practical Byzantine Fault Tolerance
2/21 Chord: A Scalable Peer-to-peer Lookup Service for Internet Applications
Pastry: Scalable, distributed object location and routing for large-scale peer-to-peer systems
2/23 Wide-area cooperative storage with CFS
Storage management and caching in PAST, a large-scale, persistent peer-to-peer storage utility [no summary required]
2/28 Handling Churn in a DHT
Designing a DHT for Low Latency and High Throughput [no summary required]
3/2 OceanStore: guest lecture by Prof. Kubiatowicz
3/7 TinyOS 3/9 TAG: a Tiny Aggregation Service for Ad-Hoc Sensor Networks
3/14 Disco: Running Commodity Operating Systems on Scalable Multiprocessors 3/16 Memory Resource Management in VMware ESX Server
Scale and Performance in the Denali Isolation Kernel[No Summary]
3/21 Debugging operating systems with time-traveling virtual machines
Speculative Execution in a Distributed File System
3/23 Reincarnating PCs with Portable SoulPads
3/28 Spring Break 3/30 Spring Break
4/4 Guest lecture: Petros Maniatis
Time, clocks, and the ordering of events in a distributed system[no summary]
Distributed Snapshots: Determining Global States of Distributed Systems
Optional reading: A probabilistic approach to distributed clock synchronization. Proceedings 9th International Conference on Distributed Computing Systems, Newport Beach, California, June 1989, pages 288-296.
4/6 Guest lecture: Petros Maniatis
Impossibility of Distributed Consensus with One Faulty Process
Reaching agreement in the presence of faults
Optional reading: The Byzantine Generals Problem
4/11 EROS: A Fast Capability System
Optional reading on KeyKos: systems: KeyKos, KeyKOS website
4/13 Labels and Event Processes in the Asbestos Operating System
4/18 Recovering Device Drivers

Improving the Reliabilitly of Commodity Operating Systems [no summary]
4/20 The SLAM Project: Debugging System Software via Static Analysis [no summary]

Automatically Validating Temporal Safety Properties of Interfaces
4/25 Solving the Starting Problem: Device Drivers as Self-Describing Artifacts [no summary] 4/27 Language support for Fast and Reliable Message-based Communication in Singularity OS
5/2 Automated Worm Fingerprinting 5/4 Secure untrusted data repository (SUNDR)
5/9 Guest Lecture by Prof. Edward Lee on how we should think about concurrency. The Problem with Threads