Instructors: Krishna Gummadi, Rodrigo Rodrigues
Meetings: Wednesdays 10:00-11:30. Wartburg building, Martin-Luther-Strasse 12.
Short description of the course:
The objective of this course is to study recent advances in the broad field of computer systems, which mainly covers areas such as distributed systems, operating systems, computer networks, or storage systems, but also interfaces to related areas like security, programming languages, computer architecture, or database systems. Towards this goal, students will read, discuss, and analyze recent papers in the top conferences of this field, particularly NSDI, OSDI, SIGCOMM, and SOSP. Students are expected to read papers in-depth prior to class, and take turns in preparing and delivering high-quality presentation of papers.
Grading: The course grades will be based on student presentations (50%), a 2-page project report and presentation (25%), and class participation (25%).
Prerequisites: Core course in Operating Systems or Distributed Systems, or equivalent background and permission of instructor.
Signing up for presentations: You must send email until Thursday, Oct 21, at 12pm with a prioritized list of at least three OSDI papers that you would like to present, and (optionally) a preferred presentation partner. You can also indicate that you prefer to wait until the second half of the course where you'll be given the option to present a networking paper.
Guidelines for presentations:
Each week we will assign one paper for discussion. The discussion
will be lead typically by two presenters. The two presenters are
expected to share the responsibilities for leading the
discussion. More specifically, they should take turns when addressing
questions from audience and they should try to split their
presentation time equally.
The time allotted for discussion is 75 mins. The discussion should
attempt to address the following questions:
1. Context for the work: What are the goals of the work, i.e.,
what problem is the paper trying to solve? Why is the problem
important or interesting? What is the state-of-the-art (i.e., related
work) in this area? What is hard about the problem?
2. Contributions of the work: What is proposed solution? What
are the key insights? Why is this the appropriate solution for this
problem? What is technically interesting or novel about the solution?
How thorough is the evaluation? Is the paper well written?
3. Potential for future work: Can you improve on the proposed
solution? What other problems in the area would be interesting and /
or hard to solve? Can the techniques behind the proposed solution be
used to solve other interesting problems?
We recommend that the presenters devote roughly equal amounts of time
(i.e., 25 minutes) for each of the above three broad categories of
questions. It is the responsibility of the presenters to manage the
discussion and cover the various questions within this time. Finally,
the audience are expected to read the paper and engage the presenters
in an active discussion.
| Date | Paper | Presenter |
| October 27 | D. Peng and F. Dabek. Large-scale Incremental Processing Using Distributed Transactions and Notifications. | T. Schaub and S. Tombers |
| November 3 | R. Geambasu, A. Levy, T. Kohno, A. Krishnamurthy, and H. Levy. Comet: An Active Distributed Key-Value Store. | A. Vahldiek and L. Galarraga |
| November 10 | J. Erickson, M. Musuvathi et al. Effective Data-Race Detection for the Kernel. | D. Wand and Zilong Wang |
| November 17 | D. Ford et al. Availability in Globally Distributed Storage Systems | Bin Cheng and W. Dai |
| November 24 | D. Ports, A. Clements, I. Zhang, S. Madden, and B. Liskov. Transactional Consistency and Automatic Management in an Application Data Cache | C. Li and P. Bhatotia |
| December 1 | W. Enck et al. TaintDroid: An Information-Flow Tracking System for Realtime Privacy Monitoring on Smartphones. | G. Doychev and E. Falk |
| December 8 | A. Aviram et al. Efficient System-Enforced Deterministic Parallelism | P. Aditya and A. Roy |
| December 15 | A. Feldman, W. Zeller, M. Freedman, and Ed Felten. SPORC: Group collaboration using untrusted cloud resources. | H. Rohdin and A. Wieder |
| January 12 | J. Pujol et al. The Little Engine(s) that could: Scaling Online Social Networks | Juhi Kulshrestha and Mainack Mondal |
| January 19 | B. Heller et al. ElasticTree: Saving Energy in Data Center Networks | Ekin Akkus and Alexey Reznichenko |
| January 26 | David R. Choffnes, Fabian E. Bustamante, Zihui Ge. Crowdsourcing service-level network even monitoring. | Farshad Kooti and Tobias Ellinghaus |
| February 2 | Project presentations (10am-12pm) | |
| February 9 | Project presentations (10am-12pm) |
February 2
February 9