The preparatory talk covers all the basics you'll need for this workshop, including:
- Linux syscalls and glibc wrappers
- chroot vs pivot_root
- namespaces
- cgroups
- capabilities
- and more
Use the provided slides while advancing through the levels, adding more features to your container. Remember to go over each level's readme, and if things get rough - you can always find the solution for level N in the level N+1 skeleton.
Not all the necessary system calls are exposed in python's standard library. In addition, we want to preserve the semantics of the system calls and use them as if we were writing C. We, therefore, wrote a python module called linux (take a look at linux.c) which exposes the relevant system calls. Have a look at the module documentation for more info.
There are currently 3 options to start the workshop by yourself:
- We created a public AMI with the required configuration and utilities
already installed:
- us-west-1: ami-07f8ee67
- us-east-1: ami-1162bc6c
- eu-central-1: ami-9a663471
- We provide a packer template so you can create your own AMI.
- We have a Vagrantfile for you to run using your favorite virtual machine hypervisor (NOTE: not yet fully tested).
The workshop material is checked out at /workshop on the instance:
/workshop/rubber-docker- this repository, where you do all the work/workshop/images- images for containers, already populated with ubuntu and busybox images
Before starting the workshop, go over the prep docs in the docs folder.
Start the workshop at /workshop/rubber-docker/levels/00_fork_exec.
This workshop has been publicly given in many places starting February 2016.
- Opstalk meetup, Tel-Aviv, February 2016
- DevOps Sydney meetup, Sydney, June 2016
- DevOpsDays Amsterdam, Amsterdam, June 2016
- SRECon EU, Dublin, July 2016
- Sela Developer Practice, Tel-Aviv, June 2016
- SRECon US, Santa Clara, March 2018
- DevOpsDays Kiel, Kiel, May 2018
Because we feel the only way to truly understand something to build it from scratch - and Linux containers are a very hyped and poorly understood technology
Of course! If you do, please consider letting us know on Twitter (@nukemberg and @nocoot) and of course send feedback.
Yes, no way we can cover the entire feature set of a real container engine. We tried to concentrate on the thing we believe are important for understanding how containers work
See contributions below
Contributions are welcome! If you found a bug or something to improve feel free to open an issue or a pull request. Please note that the entire repository is under MIT license and your contribution will be under that license.