Busy Engineers Guide to Deploying on K8s Part 2

Post Valid as of October 12, 2021.

Busy Engineer’s Guide to Deploying On K8s

This is a continuation from the last post in the series. There are three primary ways to run Kubernetes locally on a notebook: Kind, k3d, and Minikube. I personally far prefer k3d due to it having good defaults to get up and running with stateful sets long before the other two and it can create large worker nodes, these have lead to a loyalty to the product. However, my org seems to have standardized on Kind for most of their CI so I will be covering that here as it is the easiest for my coworkers to find help and documentation. It is also an excellent product to get started with Kubernetes.

Install Kind Mac and Linux

Assuming you followed the last guide we will use Homebrew again install kubernetes cli brew install kubernetes-cli
Install kind brew install kind
Then create a cluster kind create cluster
To verify this is all working run kubectl get nodes where you should see on node

Setup kind to access your GitHub credentials

First make a GitHub token that has read:packages access following the directions here
Next create a Kubernetes secret where we can store our GitHub package credentials note (change the user to your GitHub user and password to match the token) kubectl create secret docker-registry regcred --docker-server=ghcr.io --docker-username=<changeme> --docker-password=<changeme>
create a Kubernetes deployment to access your image and get it running (change the image name to match your GitHub user name) with the following deployment.yaml file

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nodejs-deployment
spec:
  selector:
    matchLabels:
      app: nodejs
  replicas: 2 # tells deployment to run 2 pods matching the template
  template:
    metadata:
      labels:
        app: nodejs
    spec:
      imagePullSecrets:
      - name: regcred
      containers:
      - name: nodejs
        image: ghcr.io/CHANGEME/app:main
        imagePullPolicy: Always
        ports:
        - containerPort: 3000

install the deployment kubectl create -f deployment.yaml
check on it’s progress with kubectl get pods --watch
when your nodes are running you can port-forward kubectl port-forward $( kubectl get pods -o name | head -n 1) 3000:3000
open your web browser to http://localhost:3000

What did we do

We installed the kubectl cli
We installed kind
Created a GitHub token with read rights to our new package
Created a deployment in Kubernetes that deployed the newly pushed package to our local kind instance
Used kubectl to make the deployment visible and opened it in a browser

Follow the next steps here