601.817 is a weekly seminar organized by the OrderLab. It covers latest advances in the research of computer systems including operating systems, distributed system, mobile and cloud computing. Students will read and discuss recent papers in top systems conferences such as OSDI, SOSP, NSDI, EuroSys, MobiSys, ASPLOS and FAST.
Each week, one student will present the paper and lead the discussion for the week. Other students MUST read the paper to be presented before the seminar. Do not come to the seminar to read the paper. This seminar is supposed to generate in-depth discussions. It is impossible to do so without reading the paper first. During the reading group discussion, there might be some questions that were left unanswered and required further investigation. In that case, the discussion lead should start a email thread to email@example.com afterwards to follow up.
The focus topics covered in the papers vary semester to semester. Example topics include fault-tolerance, reliability, verification, energy efficiency, and virtualization. The presenter decides which paper to present. In general, select the papers that are relevant to your research project first (i.e., depth-first). If you are not sure, it is a good idea to check with me first before preparing the presentation. Also, try to avoid picking papers that have already been picked in the past (the past schedules are linked on the leftside menu).
The presentation announcements are sent via the mailing list firstname.lastname@example.org. Students who registered for the course please email me to sign up for the mailing list.
Efficient and Scalable Thread-Safety Violation Detection --- Finding thousands of concurrency bugs during testing
Interruptible Tasks: Treating Memory Pressure As Interrupts for Highly Scalable Data-Parallel Programs
Effective Concurrency Testing for Distributed Systems
Finding Semantic Bugs in File Systems with an Extensible Fuzzing Framework
Automated Reasoning and Detection of Specious Configuration in Large Systems with Symbolic Execution
Pronto: Easy and Fast Persistence for Volatile Data Structures
Light-Weight Contexts: An OS Abstraction for Safety and Performance
Persistent State Machines for Recoverable In-memory Storage Systems with NVRam
Caladan: Mitigating Interference at Microsecond Timescales