Computer networks first Assignment Project 1 will be about using traceroute, parsing its output, and performing a statistical analysis of the traceroute results.
This project will require using Python to create a command line tool that automatically executes traceroute multiple times towards a target domain name or IP address specified as command line parameter. Based on multiple traceroute executions, the program will need to derive latency statistics for each hop between the traceroute client and the target machine.
To allow for repeatable tests, the program should also allow reading pre-generated traceroute output traces stored on multiple text files (one text output trace per file). Based on this pre-generated output, the program will need to compute the latency statistics as for the case of live traceroute execution.
Additional details about Project 1 will be provided in class.
Submission instructions Make a directory called LASTNAME-trstats containing a single Python3 file called trstats.py and the requirements.txt file generated by pip freeze. Create a .tar.gz archive containing the LASTNAME-trstats directory, name the resulting file as Project1-LASTNAME_FIRSTNAME.tar.gz, and submit it via eLC under this assignment.
Detail instructions Your command-line tool will need to support the following CLI arguments
usage: trstats.py [-h] [-n NUM_RUNS] [-d RUN_DELAY] [-m MAX_HOPS] -o OUTPUT -g GRAPH [-t TARGET] [--test TEST_DIR]
Run traceroute multiple times towards a given target host
optional arguments: -h, --help show this help message and exit -n NUM_RUNS Number of times traceroute will run -d RUN_DELAY Number of seconds to wait between two consecutive runs -m MAX_HOPS Number of times traceroute will run -o OUTPUT Path and name of output JSON file containing the stats -g GRAPH Path and name of output PDF file containing stats graph -t TARGET A target domain name or IP address (required if --test is absent) --test TEST_DIR Directory containing num_runs text files, each of which contains the output of a traceroute run. If present, this will override all other options and traceroute will not be invoked. Stats will be computed over the traceroute output stored in the text files
Essentially, the main task in this project is to write a Python3 wrapper around traceroute, so that you can programmatically run traceroute multiple times, read the latency statistics output by every run, and build a distribution of latency values over which to compute the required statistics. For instance, the main output of your program should be a file in JSON format that looks like this example:
[{'avg': 0.645, 'hop': 1, 'hosts': [['172.17.0.1', '(172.17.0.1)']], 'max': 2.441, 'med': 0.556, 'min': 0.013}, {'avg': 6.386, 'hop': 2, 'hosts': [['testwifi.here', '(192.168.86.1)']], 'max': 16.085, 'med': 5.385, 'min': 3.108}, {'avg': 26.045, 'hop': 3, 'hosts': [['96.120.4.5', '(96.120.4.5)']], 'max': 65.753, 'med': 20.298, 'min': 12.287}, {'avg': 26.819, 'hop': 4, 'hosts': [['96.110.205.9', '(96.110.205.9)']], 'max': 65.847, 'med': 20.51, 'min': 17.444}, ... {'avg': 168.84, 'hop': 18, 'hosts': [['124.83.228.222', '(124.83.228.222)']], 'max': 172.869, 'med': 166.869, 'min': 166.781}]
(NOTE: avg = average hop latency, max = maximum hop latency, med = median hop latency, min = minimum hop latency)
Also, the program should output a boxplot graph showing the latency distribution per each hop, similar to this one:example stats graph
Virtual environments: venv — Creation of virtual environments — Python 3.10.6 documentation PIP requirements file: Requirements File Format - pip documentation v22.2.2 (pypa.io)
Suggested plotting libraries: Box plots in Python (plotly.com) matplotlib.pyplot.boxplot — Matplotlib 3.5.3 documentation
Example for how to produce the test directory and files to be used with the --test command line option:
$ mkdir test_files; N=5; for i in $(seq $N); do traceroute -m 10 www.google.com > test_files/tr_run-$i.out; done