Syllabus and general information for MCS-378: Operating Systems (Fall 2002)
This course covers the basics of concurrent programming, operating
systems, networking, and distributed systems. There are two principal
reasons why a course on operating systems is practical to take (as
opposed to just interesting), even if you never expect to be involved
in the development of an operating system (which most of you won't):
Operating systems provide the classic application domain for certain
techniques that are useful in other contexts as well. In particular,
concurrent programming has traditionally been been an ``operating
systems'' topic, but is now quite broadly applicable. Other topics
that are no longer the sole province of operating systems include
resource allocation, security, and fault tolerance.
Although most of you won't write operating systems, you'll interface
with them. You may write applications programs that make use of the
operating system's facilities, or contribute to the design of hardware
on which the operating system runs, or find yourself educating an
operating system developer about the features that would allow your
web server or database system to run most efficiently.
You may be a system administrator who has to configure and tune an
understanding of what goes on inside an operating system will allow
you to be more effective in roles like these. You may even find
yourself tacking a new feature on in some small corner of an operating
system other people wrote.
I will be available in my office (OHS 303) from 11:30-12:20 on
Mondays, Wednesdays, and Fridays; from 1:30-2:20 on Tuesdays and
Thursdays; and by appointment. Or try your luck: just stop by and see
whether my door is open. You may send me electronic mail at firstname.lastname@example.org or call me at
extension 7466. I'll try to put any updates to my office hours on my
web page, so check there if in doubt.
World Wide Web
All course materials will be available through my World Wide Web page.
The URL for this course is http://www.gustavus.edu/~max/courses/F2002/MCS-378/.
After this syllabus I will give hardcopy handouts only to those
students who want them.
Text and readings
The primary text for the course will be Siberschatz, Galvin, and Gagne's
Applied Operating System Concepts, first edition, but I'll also
be distributing some readings that are recent conference papers, in order to provide both an update
on some recent interesting work and also a bit more ``meat.'
Those readings will form the basis for the student
presentations, described below. Lab handouts may also
include some non-trivial reading.
There will be four lab assignments. The due dates for the labs are
shown in the syllabus below.
We'll meet in the OHS 326 or 329 lab
every Monday, and Friday November 22nd as well.
Attendance is expected for all lab days. (If you turn in a lab
report early, you are excused from the remaining days devoted to that
lab.) I will excuse up to two absences per student, for any reason.
Use yours wisely. If you exceed this allowance, I may reduce your
course grade by one letter grade.
For each lab, I will assign you a lab group. You will work with
different people each lab, and with a different person for the
presentation than for any of the labs. Thus, over the semester you
will work with the majority of the class.
There will be twelve days in the semester when I sit back and relax and
let students teach the class instead. Each time, a pair of students will have
a class period to lead discussion on one of the papers we read from a
conference. The whole class will be expected to read the
paper at least casually, but the pair leading the class will be
expected to have read it more carefully and done whatever it takes to
really understand it, which may involve additional background reading,
talking with me, etc.
Each pair of students will be responsible for setting up a meeting
with me the week before their presentation to go over it with me.
(You are welcome to additional consultations as well.) At that
meeting, you should also give me three straightforward
questions about your reading. I will distribute these to the full
class. All class members will turn in their answers to these three
questions at the beginning of class the day of your presentation, as a
way of making sure we all have a common starting point.
I will provide
a list of attributes that good presentations have.
For each presentation, each student in the audience will be asked to
select two of these attributes that were particular strong points, and
two that particularly need work. I will summarize this feedback for
the presenters. I will also use the same list of attributes to
structure my own evaluation of the presentation, which will take place
in a face-to-face "debriefing" with the presenters, with the grade
generated as a summary of that meeting.
I'm going to ask you to express a preference ranking for the
topics, and will try to the extent possible to assign people to topics
based on that. (You can also take into account in your preference
rankings the timing of the presentations.)
I'll need to have you
email me the rankings by 1:15pm Thursday (September 5) so that I can
make the presentation and lab group assignments in time for the lab
The topics are:
"Scalable Linux Scheduling",
Stephen Molloy and Peter Honeyman,
FREENIX Track: 2001 USENIX Annual Technical Conference, pp. 285-295.
"PMQS : Scalable
Linux Scheduling for High End Servers", Hubertus Franke, Shailabh
Nagar, Mike Kravetz, and Rajan Ravindran, Fifth Annual Linux
Showcase and Conference.
October 4 (rescheduled to October 11):
Meets Z: Verifying Correctness In The Presence Of POSIX Threads",
Bart Massey and Robert Bauer,
FREENIX Track: 2002 USENIX Annual Technical Conference, pp. 221-234.
"The JX Operating System",
Michael Golm, Meik Felser, Christian Wawersich, and Jürgen Kleinöder,
General Track: 2002 USENIX Annual Technical Conference, pp. 45-58.
"The Challenges of Integrating the Unix and Mac OS Environments",
2000 USENIX Annual Technical Conference, invited talk,
updated version as of May 5, 2002.
"GPFS: A Shared-Disk File System for Large Computing Clusters",
Frank Schmuck and Roger Haskin,
Proceedings of the Conference on File and Storage
Technologies, 2002, pp. 231-244.
"Track-Aligned Extents: Matching Access Patterns to Disk Drive Characteristics",
Jiri Schindler, John Linwood Griffin, Christopher R. Lumb, and Gregory R. Ganger,
Proceedings of the Conference on File and Storage
Technologies, 2002, pp. 259-274.
"Ningaui: A Linux Cluster for
Andrew Hume and Scott Daniels,
FREENIX Track: 2002 USENIX Annual Technical Conference, pp. 195-206.
"Structure and Performance of
the Direct Access File System
Kostas Magoutis, Salimah
Addetia, Alexandra Fedorova,
Margo Seltzer, Jeff
Chase, Drew Gallatin, Richard
Wickremesinghe, and Eran Gabber,
General Track: 2002 USENIX Annual Technical Conference, pp. 1-14.
Hao Chen, David Wagner, and Drew Dean,
11th USENIX Security Symposium, August 2002.
"Deanonymizing Users of the SafeWeb Anonymizing Service",
David Martin and Andrew Schulman,
11th USENIX Security Symposium, August 2002.
"How to 0wn the Internet in Your Spare Time",
Stuart Staniford, Vern Paxson, and Nicholas Weaver,
11th USENIX Security Symposium, August 2002.
Homework assignment policy
I will assign homework problems.
You may turn in any individual homework problem whenever you think you have it
solved. I will return it to you as quickly as I can, but normally
with only an indication of whether it is acceptable or needs more
work. (Sometimes I may give a brief indication of what area it needs
more work in.) If a problem needs more work, and you aren't sure what
sort of work it still needs, you should treat that as an invitation to
come talk with me about it. Once you've done the additional work, you
may turn the problem in again. In fact, you may turn in each problem
in as many times as you like, until it is marked as acceptable. Your
grade for the homework portion of the course will be based on the
fraction of homework problems that you eventually did acceptably.
Normally homework problems may be turned in at any time up until 10:30
on October 16 for chapters 1-10, and 10:30 on December 13, for the
remainder. However, if we would benefit from discussing a homework
problem in class, I may issue a "last call" for solutions to that
problem, at least a week in advance.
Unless I indicate that a particular problem must be done individually,
you may work on any problem in a group of two or three students.
One copy of the solution produced by the team should
be turned in, with all team members names on it. Write "we all
contributed fairly to this solution" and have all team members sign
under that statement.
There will be midterm and final exams.
Based on experience in past years, I have switched the midterm exam
to the evening, in order to provide a less time-pressured test
format. As shown in the syllabus, the midterm will be 7:00-8:30pm on
October 22 in OHS 317. Please let me
know as soon as possible if you won't be able to take the test at
that time. In that case, I will be happy to set up another
hour and a half time block that works for you.
The final exam will be as scheduled by the
Students are encouraged to discuss the course, including issues raised
by the assignments. However, the solutions to assignments
should be individual original work unless otherwise specified. If an
assignment makes you realize you don't understand the material, ask a
fellow student a question designed to improve your understanding,
not one designed to get the assignment done. To do otherwise is to
cheat yourself out of understanding, as well as to be intolerably
Any substantive contribution to your solution by another person or
taken from a publication should be properly acknowledged in writing.
Failure to do so is plagiarism and will necessitate disciplinary
The same standards regarding plagiarism apply to team projects as to
the work of individuals, except that the author is now the entire team
rather than an individual. Anything taken from a source outside the
team should be be properly cited.
One additional issue that arises from the team authorship of
project reports is that all team members must stand behind all reports
bearing their names. All team members have quality assurance
responsibility for the entire project. If there is irreconcilable
disagreement within the team it is necessary to indicate as much in
the reports; this can be in the form of a ``minority opinion'' or
``dissenting opinion'' section where appropriate.
Late lab assignments
All lab assignments are due at the beginning of class on
the day indicated. Late assignments will be penalized by one ``grade
notch'' (such as A to A- or A- to B+) for each weekday late or fraction
thereof. However, no late assignments will be accepted after graded
assignments are handed back.
If you are too sick to complete an assignment on time, you
will not be penalized. Simply write ``late due to illness'' at the
top of the assignment, sign your name and hand it in. Other circumstances
will be evaluated on a case-by-case basis.
Please point out any arithmetic or clerical error I make in grading,
and I will gladly fix it. You may also request reconsideration if I
have been especially unjust.
I will provide you with a letter grade on each lab
assignment, on your presentation, and on each test, in addition to the mid-term and final
grades, so that you may keep track of your performance. As a
guideline, the course components will contribute to your final grade
in the proportions indicated below:
- 40% Labs (4 @ 10% each)
- 10% Presentation
- 18% Homework (based on fraction done; see above)
- 12% Midterm
- 20% Final exam
All homework and lab reports should be readily readable, and should
not presuppose that I already know what you are trying to say. Use
full English sentences where appropriate (namely almost everywhere)
and clear diagrams, programs, etc. Remember that your goal is to
communicate clearly, and that the appearance of these technical items
plays a role in this communication process. Be sure your assignments
are always stapled together and that your name is always on them.
Please contact me immediately if you have a learning or physical
disability requiring accommodation.
In the reading column, a single number with no decimal point indicates
an entire chapter. Section 0 means the material at the beginning
of a chapter before the first section.
This is my best guess as to the rate at which we will cover material.
However, don't be shocked if I have to pass out one or more revised
|9/9||Lab 1: Scheduling experiments||
|9/16||Lab 1 (concludes)||
|9/18||paper 1||Scalable Linux Scheduling||
|9/20||paper 2||Pooled Multi-Queue Scheduling||
|9/23||Lab 2: Kernel-level experimentation||Lab 1
|9/25||7.7-7.10||More on process synchronization||
|9/30||Lab 2 (continued)||
|10/4||Lab 2 (continued)||
|10/7||Lab 2 (concludes)||
|10/11||paper 3||Verifying Correctness||
|10/14||10.5-10.8||More on virtual memory||
|10/15||paper 4||The JX Operating System||
|10/16||review/catch-up||HW, chapters 1-10
|10/17||no class, just a due date (at 5pm)||Lab 2
|10/22||mid-term exam, 7:00-8:30pm, OHS 317||
|10/25||11.6-11.11||More on file systems||
|10/28||Lab 3: Filesystem locality||
|10/29||paper 5||Integrating Unix and Mac OS||
|10/30||paper 6||A large-cluster file system||
|11/4||Lab 3 (continued)||
|11/6||paper 7||Track-Aligned Extents||
|11/11||Lab 3 (continued)||
|11/12||More on networking||
|11/18||Lab 3 (concludes)||
|11/19||paper 8||A Linux Cluster||
|11/20||17||Distributed file systems||Lab 3
|11/22||Lab 4: Distributed communication (a Friday lab)||
|11/25||Lab 4 (continued)||
|11/26||paper 9||Direct Access File System||
|12/2||Lab 4 (continued)||
|12/3||paper 10||Setuid Demystified||
|12/4||Lab 4 (continued)||
|12/6||paper 11||Deanonymizing Users of SafeWeb||
|12/9||Lab 4 (concludes)||
|12/10||paper 12||How to 0wn the Internet||
|12/13||review/catch-up/evaluation||HW and Lab 4
Course web site: http://www.gustavus.edu/~max/courses/F2002/MCS-378/
Instructor: Max Hailperin <email@example.com>