CS 2510 - Advanced Operating Systems

• Intro to Distributed Systems (size: 515K)
• Intro to Clocks and Clock Synchronization (size: 175K)
• Distributed Mutual Exclusion (size: 860K)
• Distributed Mutual Exclusion Notes from Xin Yuan, FSU (Pitt Alumnus!)
• Consistency (size: 112K)
• Replication (size: 245K)
• Load Balancing (size: 175K)

Readings, Textbooks, and Prelims

To complete the PhD preliminary exam requirements, a student must take both the midterm and final exams of the course, and answer the additional question(s) that are designated as such. These questions draw from material that is expected from the undergraduate operating systems course. The midterm will include at least one additional question focusing on uniprocessor synchronization problems, while the final is expected to include at least one question dealing with file systems and/or I/O scheduling. Either the midterm or final can include questions on scheduling algorithms. In addition to the material covered explicitly in this course, the reading list for prelim preparation includes the following from the undergraduate text:

Grading

Note: You must average above 50% in each of the following two parts in order to pass the class: (1) the programming assignments and  (2) the exams.  Lower than 50% on either portion of the class will result in a failing grade regardless of the overall score.

• Intro to advanced/dist operating systems
• Distributed Systems: architecture, motivation, issues
• Communication Models
  • IPC, RPC, message passing, shared memory
  • Client-Server Computing: paradigm, examples, discussion
• Threads and processes
• Distributed synchronization
  • introduction, definition, models for centralized and distributed, causal/total ordering
  • implementation: lamport, ricart&agrawala, maekawa, vector clocks, etc
• Distributed ME:
  • introduction, requirements, (non- and) token-based algorithms
  • generalized non-token-based algorithm
  • token-based algorithm and comparisons
• Agreement Protocols
• Distributed resource management
  • load balancing, load sharing
  • resource allocation in real-time systems
• Principles of Fault-Tolerance (interspersed with other topics)
  • coding theory (introduction along communication models)
  • redundancy (introduction along scheduling and agreement protocols)
• Principles of Security and Protection
• Distributed shared memory (time permitting)
• Distributed file systems (time permitting)

Attendance

Class attendance is officially mandatory. Homeworks, assignments, and important dates will be posted on the class web page, but this is provided as a courtesy and is not always complete. It is your responsibility to keep up with class materials and assignment.

Office hours are optional. They are your chance to ask question and get help from the professor and the TA about the material being covered, the programming assignments, etc.

Academic Honesty: Collaboration vs. Cheating copied from Ahmed Amer

This really should not be an issue, but to make things as clear as possible the following is necessary.

You are encouraged to discuss the course material and concepts with other students in the class. However, all work that you submit must be your own. Under no circumstances may you look at anyone else's code or show anyone else your code. And while you may discuss the concepts used in the programming assignments, you may not discuss implementation details of the assignments themselves.

If you are caught copying or otherwise turning in work that is not solely your own, you will fail the course.

The bottom line is that you are expected to conduct yourself as a person of integrity - you are expected to adhere to the highest standards of academic integrity. This means that plagiarism¹ in any form is completely unacceptable. As a (soon to be) computing professional, I encourage you to consult the code of ethics appropriate to your discipline².

Plagiarism will be assumed until disproved on work that is essentially the same as that of other students. This includes identically incorrect, off-the-wall, and highly unusual duplicate answers where the probability of a sheer coincidence is extremely unlikely. All parties to this unacceptable collaboration will receive the same treatment.

You should bring a picture identification with you to all examinations and be prepared to show it upon request.

If you are unsure of what is and is not allowed by this policy, talk to the instructor.

¹ pla-gia-rize vt. to steal and pass of as one's own (the ideas or words of another) to present as one's own an idea or product derived from an existing source - pla-gia-riz-ern. (source: Webster's New World Dictionary).
² The Association for Computing Machinery is http://www.acm.org/, the IEEE is http://www.ieee.org/ and the IEEE Computer Society is http://www.computer.org/.

OLD SAMPLE 2nd midterm study guide here . Note that this is just set of things that you should know. you're actually responsible for all the stuff in the reading list (book chapters, classnotes, handouts, and class discussions).