First contact

Kubernetes is an orchestrator for deploying containers but you can think about it as an Operating System running your programs. While it's good to know how an Operating System works (and you should keep learning Kubernetes internals beyond this course), as long as you know how to install, upgrade and verify your programs are running fine, that'll be enough.

You will always see Kubernetes as a single component but under the hood it runs on a Cluster of machines called Nodes. More nodes means more availability. Bigger nodes means more memory and CPU available to your programs.

Every Object managed by Kubernetes is represented by a RESTful resource and we'll see how, at the end of the day, kubectl is just a smart HTTP client.

For the scope of this course, we won't need a fully functional multi-node Kubernetes cluster. Instead, we'll be using Kind. KIND stands for Kubernetes IN Docker and it's capable of managing local clusters of one node orchestrating a bunch of Docker containers. At this stage, feel free to ignore the internals and look at Kind for what it is, a tool to start/stop Kubernetes clusters.

Another tool you must get familiar with is kubectl, the CLI tool you'll be using to do stuff on any Kubernetes cluster, whatever it's local or a production system. One caveat about kubectl, you need to double check the version you have in your system is compatible with the version of Kubernetes that runs in the cluster: client and server can be at most 1 minor version distant from each other. In our case, the versions are pinned so as long as you've followed the setup instructions you should be good.

In this unit we'll start a local Cluster and we'll explore its components with kubectl.

Exercise n.1: start the Cluster

If kind is correctly installed, all you have to do is running this command from the root of the repo:

$ kind create cluster --name k8s101 --image kindest/node:v1.18.2 --config kind-config.yaml
Creating cluster "kind" ...
 ✓ Ensuring node image (kindest/node:v1.18.2) 🖼
 ✓ Preparing nodes 📦
 ✓ Writing configuration 📜
 ✓ Starting control-plane 🕹️
 ✓ Installing CNI 🔌
 ✓ Installing StorageClass 💾
Set kubectl context to "kind-kind"
You can now use your cluster with:

kubectl cluster-info --context kind-kind

Have a nice day! 👋

Notice we use a specific image (and not latest) so that we know it'll work with the pinned kubectl version. If you see no errors the default Cluster called kind should be ready to use, and you can confirm it by running:

$ kind get clusters
kind

Exercise n.2: explore the Cluster

To get informations about the Cluster, just run

kubectl cluster-info

To prove that everything is a RESTful resource in Kubernetes, you can increase kubectl's log verbosity and see how it performs HTTP calls under the hood:

kubectl cluster-info -v6

We can query the status of few key components of the Cluster:

kubectl get componentstatuses

And we can also get the list of all the nodes composing the Cluster:

kubectl get nodes

The more detailed, more structured version of kubectl get is the command describe, that gives us a comprehensive view of the Objects we query, in this case the nodes of the Cluster:

kubectl describe nodes