Part I.
You are highly encouraged to co-operate. But, I will expect
that everyone will write (or type) his/her own answers.
This is not a joke: you will learn more from working with others.
Do the following questions and problems from the
reference book. If you do not have the book, come and see me.
You are expected to do all the questions, but you have to do
only 80% of the problems.
Please answer briefly, but with complete sentences that can make sense
even to someone who does not know the question.
Most answers should be one paragraph or so.
The official language of the class is english.
Chapter 3
Questions:
3.1-3.4: 1, 2, 3
3.7:
8
Problems: 1, 2, 10, 11, 12, 14, 16
Chapter 4
Questions
4.1-4.4: 1, 3, 4
4.5 : 7, 10, 12, 15
4.8 : 20, 23, 24
Problems
1, 2, 11, 14, 19, 20, 21
Part II: Measure the Internet Performance
Using traceroute, probe 1,000 times a path of your choice with at least
more than 10 hops. Then examine the statistical properties of the path.
For example:
what is the stability of the path? do you always get the same path?
a) what is the path-length distribution (avg, standard deviation,
min, max)
b) what is the round-trip-delay distribution
c) how many times did traceroute failed to reach the destination
Comment and try to explain the phenomena you observe.
It will also be interesting to plot some of your results
(i.e. distribution of path-legth) using gnuplot or some other
plotting tool. These are the minimum requirements. You are highly
encourage to examine more aspects.
For example, you can see if there is a periodicity in the
routing phenomena you observe.
See Paxson, V., End-to-End Routing Behavior in the Internet, on main
page, for some ideas.
Instead of traceroute, you can use pathchar with 100 repetitions since
it takes longer. Actually, a neat idea is to compare the results of
traceroute and pathchar
You are encouraged to explore other aspects of path stability.
It also makes sense to write a script in your favourite shell and use
awk to >manipulate the results.
Here is my awk file as a shell script for finding averages and confidence intervals:
awk-filter
USAGE: nzavg.awk column-no datafile
Here is a
script for changing a string in a files and outputing the new file
Usage : sub oldPattern newPattern files