Don’t stop now! The next scenario will only take about 10 minutes to complete.

Kubernetes Playground

Step 1 of 8

Introducing kubectl

Here we have a Kubernetes cluster with two nodes: one master node and one worker node. Both nodes have kubectl installed.

kubectl is the command line tool that we use to communicate with a Kubernetes cluster. You can view Kubernetes resources by running kubectl get <resource-type>. Let's take a look at the nodes in our cluster:

kubectl get nodes

You can see we have two nodes, one master and one worker.

Let's look at the components of the cluster:

kubectl get componentstatus

This shows us that etcd, the scheduler, and the controller-manager are all working correctly.

Let's see if there are any pods:

kubectl get pods

We haven't deployed any pods yet, so that's expected.

In fact Kubernetes is already running some Pods under the hood. You can split up a Kubernetes cluster into different namespaces. So you can give different teams namespaces and they won't clash with each other.

If we look at Kubernetes' internal namespace, kube-system, by adding -n kube-system we can see that Kubernetes is actually running several of its own components in pods. That's a cool thing about Kubernetes, once you've bootstrapped some of the components, it can run the rest of the components itself.

kubectl get pods -n kube-system -o wide

Terminal Host 2