Summary

By now, you should have a basic understanding of the stack of software that makes a modern container orchestrator like Kubernetes tick.

You should now understand the following:

• Containers are built on top of much lower-level features in the Linux kernel, such as namespaces and Cgroups.
• In Kubernetes a pod is a powerful abstraction that is built on top of containers.
• Kubernetes uses control loops to build a powerful system that allows the operator to declaratively specify what should be running. Kubernetes automatically takes actions to drive the system towards this state. This is the source of Kubernetes' self-healing properties.
• Nearly everything in Kubernetes can be given a label, and you should label your resources in order to make managing them simpler.

In the next chapter, you will gain some practical experience using the Kubernetes APIs by running a small cluster on your workstation.