/CVE-2022-4096

This experiment is destinated to demonstrate how the DNS rebinding attack works on an emulated IoT. In the setup, we have a simulated IoT device, which can be controlled through a web interface (this is typical for many IoT devices).

Primary LanguageCSS

CVE 2022 -4096

Amine TITROFINE | January 21, 2023


This experiment is destinated to demonstrate how the DNS rebinding attack works on an emulated IoT. In the setup, we have a simulated IoT device, which can be controlled through a web interface (this is typical for many IoT devices). Many IoT devices do not have a strong protection mechanism, if attackers can directly interact with them, they can easily compromise these devices.

Environnement

Host Machine : - This exploit has been experimented on (Linux kali 6.0.0-kali5-amd64), it can also be tested on (Ubuntu) distributions

Exploit

Containers commands and setup

First, clone this project in your local machine

$ git clone https://gitlab.com/grenoble-inp-ensimag/Secu3A/Devoir2/CVE_2022_4096_amine_titrofine_farah_ben_youssef_walid_lanjri.git

we access to the directory that contains the files of our repositroy

$ cd CVE_2022_4096_amine_titrofine_farah_ben_youssef_walid_lanjri

we start by building all the defined services in the (docker-compose.yaml) file

$ docker-compose build

And then, we ran the following command to start the different services

$ docker-compose up

Configure the User VM

(Step 1. Reduce Firefox’s DNS caching time:)

network.dnsCacheExpiration: change its value to 0 (default is 60)

(Step 2. Change /etc/hosts:)

192.168.60.80 www.seedIoT32.com

(Step 3. Local DNS Server:) we add the nameserver entry in the resolver configuration file (/etc/resolv.conf).

nameserver 10.9.0.53

Testing the Lab Setup.

After configuring the User VM, use the dig command to get the IP address of www.attacker32.com. You should get 10.9.0.180 . If you do not get this, your lab environment is not set up correctly.

$ dig http://www.attacker32.com

Launch the Attack on the IoT Device

This part is well documented in the report, please refer to it starting from page (17).