Skip to content

Latest commit

 

History

History

backend

README.md

Dependencies

xhost: apt install x11-xserver-utils

Architecture

The overall architecture of the test suite is given in the block diagram below.

  • Networking is in yellow
  • Applications are in red
  • Docker is in blue

Architecture

HTTP Modification Demo

1. Launch Terminals

Three interactive shells are required for a full demonstration.

  • Leftmost shell will be for a gui_instance container
  • Middle shell will be for a lens_instance container
  • Rightmost shell will be for server_instance container

2. Start Test Harness

Start the test harness using the testharness.sh wrapper script.

sudo ./testharness.sh http

3. Confirm that Containers Have started

Confirm that all three containers listed below are running.

  • viewer_instance
  • camera_instance
  • lens_instance

4. Attach to Other Containers

The middle shell should already be attached to a lens_instance container.

In the leftmost shell, attach to gui_instance.

sudo docker container attach gui_instance

In the rightmost shell, attach to server_instance.

sudo docker container attach server_instance

5. Load the Website

Within the leftmost shell inside the container, load a website from the server.

firefox -private 192.168.0.2

Observe that modification should have occured. To confirm, place the driver back into passive mode to confirm. Within the middle shell inside the container run:

driver passive

Automated Testing

Start the test harness using the automated test contol script test.py. To save time the script can either generate results and save them to a temp file with

sudo ./test.py --save

Or load results from a temp file and generate the associated plots.

sudo ./test.py --load

Other Commands

Raw capture with TCPdump

To generate packets on a specific interface, either cURL or ping can be used.

curl --interface eth0 localhost:80
ping -I eth0 localhost:80

To capture packets on an interface, use TCPdump.

tcpdump -e -n -i -vvv -xx eth0
sudo tcpdump -e -n -vvv -xx -w cap.hex -i eth0

Hexdump can be used to load captures which were previously saved.

hexdump -C cap.hex

Download a website

Use wget to pull a servable copy of the website recursively.

wget -k -H -E -p -nd --no-parent -e robots=off https://en.wikipedia.org/wiki/Man-in-the-middle_attack

Likely this will result in file names/paths which contain characters like ? which are invalid on windows and should not be commited to git. Therefore zip the resulting website by navigating into the folder and running:

tar -czvf ../payload.tar *

Extra shells

When attaching multiple times to a docker container, all connections share the same stdin and stdout. To get a second shell to a container, run one of the commands below.

sudo docker exec -it iperf_instance bash
sudo docker exec -it gui_instance bash
sudo docker exec -it lens_instance bash
sudo docker exec -it server_instance bash

Wireshare

Install wireshark

sudo apt update
sudo apt install wireshark

Note that wireshark required root permissions to capture packets in its default configuration and so should be run with sudo wireshark from a new shell.

If there are both public and private ip addresses in the capture, then consider only ones in the ranges of private IPs. 1) 10.0.0.0 to 10.255.255.255 2) 172.16.0.0 to 172.31.255.255 3) 192.168.0.0 to 192.168.255.255

Other

sudo apt-get install python3-matplotlib

https://stackoverflow.com/questions/26999595/what-steps-are-needed-to-stream-rtsp-from-ffmpeg

The docker daemon is only accessible to root users by default. We can fix this with the following commands.

sudo groupadd docker
sudo usermod -aG docker $USER

See https://docs.docker.com/engine/install/linux-postinstall/ for more