DIY Container

These scripts are designed to highlight the basic anatomy of a container. They provide basic network, IPC, PID, UTS, cgroup, and mount namespacing, as well as filesystem isolation.

Pre-requisites

• A bash shell
• 12Mi disk space

Usage

The following commands require root privileges. The container filesystem behaves like a "real" container, thus any changes made within runtime-exec will not be persisted. You can run either script with VERBOSE=1 ./<script> <args> to see what commands are being executed.

1. cd to the repo directory
2. Execute ./setup-image to setup the container image (uses alpine)
3. Execute ./runtime-exec <cmd>, where <cmd> is the command to execute within the container image (i.e. sh)
   1. The container image contains httpd, which can be run by passing httpd -D FOREGROUND as the command and connecting to http://10.178.61.2
   2. You can examine the network and PID containment with ip, ping, and ps

TL;DR

git clone https://github.com/Nuvalence/diy-container.git
cd diy-container
sudo ./setup-image
sudo ./runtime-exec httpd -D FOREGROUND

Assuming all went well, the containerized httpd instance should now be accessible at http://10.178.61.2.