NEU Logo Information Storage Technologies
CS G389 Special Topics in Operating Systems

Fall Semester 2006
Home | Course Schedule
Students are responsible for all materials appearing on the Web site.
Last Updated:

News

The lecture notes from lecture 11 are now online.

The readings for next week are posted. There are three papers. The presenters for this week are:
HQ Replication - Da Lin
Rethink the sync- Sapan Raval
Pastiche - Neeraj Joshi

Old News

The lecture notes from Jim Espy's lecture are on-line now. Make sure that you hand in your proposals by Wednesday, 11/15 night.

The readings for next week are posted. There are two papers. And the presenters for this week are (thorough pseudo-random process):
Elephant - Mike Rieker
Ext3cow - Xin Dong

L7 presenter:
Logical vs. Physical Backup - Nelson Seto

Each person presenting a paper, summarize the paper in your own words and prepare to lead the discussion. To do so, prepare two or three questions that you want the class to discuss. Each presenter should think of leading a discussion for about 1/2 hour. Assignments for discussion in Lecture 6:
MapReduce - Chintan Shah
Bigtable - Dan Kunkle

The lecture notes (slides) PDF materials have been posted. Checkout the couse schedule page for the links. Note that you can access the PDFs only if you have an NEU/CCS account.

This course is be substantially different from the 2005 offering, focusing more on the distributed nature of systems and the challenges in managing them.

Instructor
Jiri Schindler
NetApp Inc.
email: jiri@ccs.neu.edu

Office Hours
By appointment (via e-mail).

Class Location & Time
Room TBD
Tuesdays 6pm - 9pm

About the instructor
Jiri Schindler is a member of the research staff at NetApp, the makers of enterprise-scale network-attached storage systems. Prior to joining NetApp, he was a Principal System Architect at EMC's Centera division, designing content-addressable clustered object-stores. He received his PhD from Carnegie Mellon University.

Syllabus

Description
Access to information is arguably the most precious asset in today's society. Mining large application-specific databases to quickly produce relevant information resulted in unprecedented commercial success of Google's Page Rank algorithm or Amazon's "Customers who bought this item also bought..." feature. Systems like these employ thousands of computers that execute tens of distributed and specialized applications.

Secured storage and timely access to information is an active area of computer science research, spanning, among others, fields of networking, databases, storage, distributed systems theory, security, and dependability. This course stresses the inter-disciplinary nature of building and understanding information storage systems. It covers the underlying technologies, their trends, and practical aspects of algorithms in distributed computer systems. The topics include architecture of enterprise-scale storage systems, database organization, performance analysis, management and disaster recovery, information security and retention in the face of recent regulatory changes (Sarbanes-Oxley, SEC 17-4a, HIPPA etc.). The course will study many examples of real-world systems and draw from current research literature.

Prerequisites
For students in Software Engineering concentration, CS G110 or CS G270.
Because of its interdisciplinary nature, the course is targeting motivated students enrolled in Master's or PhD programs. Highly motivated undergraduates (seniors) can enroll if they provide a recommendation of their academic or research project advisor. Students must have basic knowledge of networking, software development, and have taken fundamentals of operating systems or equivalent, As we will be reading many primary sources, it is expected that students know how to read and critique technical papers. Successful completion of any (one) of these courses (or their more advanced counterparts) is also suitable: CS G130 Introduction to Database Systems, CS G150 Fundamentals of Computer Networking, CS G112 Computer Systems, or CS G389 Enterprise Storage Systems.

Expectations
This is a first offering of this course. It is designed for highly motivated students. Students should expect for the material to be unpolished and evolving through out the term. In return, the instructor promises to work hard to make this as fun and academically enriching experience as possible.

To get the most out of the lectures and to effectively participate in the class discussions, all materials assigned for a given lecture must be read beforehand.

Text and Other Course Materials
There is no single textbook for this course. Handouts, book chapters, articles, and URLs will supplement course material. These will be made available online when possible and will be posted on the couse website.

Grading
Grades will be determined based on the following criteria.

Midterm Examination: 20%
Final Examination: 30%
Class Participation: 20%
Homeworks & Projects: 30% 

Exams will be closed book.
All assignments must be completed successfully to fulfill the course requirements.

Homework and Projects
Homeworks will be written assignments that take one week to complete.

Projects involve implementation and a written report.  Two or three projects will be assigned throughout the semester.  Each should take two weeks to complete.

Academic Conduct
Unless otherwise indicated, students may only discuss concepts for the homework with each other. Submissions should represent the student's individual effort. On the handed assignment, clearly indicate the person(s) with whom you discussed the concepts.

Page last updated: