To run our first pod, we'll first need a kubernetes cluster. We'll use minikube, which is essentially a single node cluster. to get minikube installed on your machine, follow the instructions here
To start a minikube machine, run the following from a terminal -
$ minikube start --memory 4096You can find additional information in the official docs
Once minikube finishes the bootstrapping process, we can start interacting with the cluster (through the API server) using kubectl. Note that minikube wrote an entry the kubeconfig file and set the context to minikube itself.
Let's explore what components are in out cluster -
$ kubectl get podsYou'll get nothing. Why? because we haven't deployed any resources to our cluster. However, we can list the control plane components which are found in the kube-system namespace -
$ kubectl get pods --namespace kube-systemThere, better.
Now, to run our first pod -
$ kubectl run nginx --image nginx:alpine --port 80This is very similar to a docker run command. We tell kubectl to create a deployment called nginx with the nginx:alpine image and use container port 80. While we haven't told kubectl to create a deployment explicitly, it does so by default since we haven't defined the restart policy.
To see our new pod starting, we can run -
$ kubectl get podsAnd see it's status. From here we can use other commands to get information about the pod such as kubectl describe and kubectl logs. To scale the application we could run -
$ kubectl scale deployment nginx --replicas=3And see other pods start as well.
Our pods are running but are not yet accessible from outside our cluster. To expose them, we use -
$ kubectl expose deployment nginx --type NodePort --name nginxThis will create a service of type NodePort. to see a list of services we run kubectl get services. note the external port for out service (the high one, above 30000). We can than get the minikube VM ip using minikube ip and point our browser to http://<minikube-ip>:<service-port>