I have discovered a ZipSlip vulnerability in a web application. Simply put, ZipSlip is basically a directory traversal through a specially crafted zip file, which is made to contain "../" inside, in order to extract to an arbitrary directory.
I have deployed the vulnerable application using Docker. The attack is executed successfully, the file can be extracted several directories below the original one. However, I cannot access the root directory of the docker image.
Let's assume the following directory tree: root -> folder1 -> folder2 -> folder3
The file is supposed to extract in folder3. Using my malicious zip I can extract the contents in folder2 and folder1. But if I put more "../" to reach root and then some other folder in the root level, I cannot.
I'm trying to figure out what is the measure in place that prevents me from reaching the root directory.