A while back I created a Go container which returns the hostname of the container its running on - markkerry.github.io - Running a Local Go Web App in Docker. Running this container from a Kubernetes cluster was what I had in mind when I created it. In this post I will walk through the steps I completed to create the Kubernetes deployment with an image pulled from Docker Hub.

I will be using the same Ubuntu server lab running on VirtualBox - markkerry.github.io - Ubuntu Server 20.04 LTS Lab. And the same Kubernetes cluster built from this post - markkerry.github.io - Running a Kubernetes Cluster on Ubuntu and VirtualBox

The official Kubernetes docs is what I used for reference - kubernetes.io - Pull an Image from a Private Registry.

Pull an Image from Docker Hub

From the Control Plan server I logged into Docker Hub.

docker login

This created ~/.docker/config.json

Then I had to copy the credential into Kubernetes:

kubectl create secret generic regcred \
    --from-file=.dockerconfigjson=/home/mark/.docker/config.json \
    --type=kubernetes.io/dockerconfigjson

(Optional) View the secret in YAML format:

kubectl get secret regcred --output=yaml

Create Deployment

Next I created a yaml file for the deployment of the Go app and specified the docker id:

touch ~/go-app-pod.yaml

I edited the file with the contents below:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: go-app-deployment
  labels:
    app: go-app
spec:
  replicas: 2
  selector:
    matchLabels:
      app: go-app
  template:
    metadata:
      labels:
        app: go-app
    spec:
      containers:
      - name: go-app-pod
        image: markkerry/go-app:v1
        ports:
        - containerPort: 3000

I only specified 2 replicas. Note the image is my Docker ID and the app.

Apply Deployment

After saving the file it was time to apply to the pods:

kubectl apply -f go-app-pod.yaml

kubectl get deployments

kubectl get pod

NAME                                READY   STATUS    RESTARTS        AGE
go-app-deployment-f687dd5bc-824tw   1/1     Running   0               13s
go-app-deployment-f687dd5bc-v7c56   1/1     Running   0               13s

As below:

allcommands

Test Deployments

The following command will show the IP of each pod and the node it is running on:

kubectl get pod -o wide

NAME                                READY   STATUS    RESTARTS        AGE     IP              NODE        NOMINATED NODE   READINESS GATES
go-app-deployment-f687dd5bc-824tw   1/1     Running   0               21s     10.10.166.135   node1       <none>           <none>
go-app-deployment-f687dd5bc-v7c56   1/1     Running   0               21s     10.10.104.5     node2       <none>           <none>

And finally can curl each one which will return the host name of the pod:

curl 10.10.166.135:3000
curl 10.10.104.5:3000

curlPods