Cloud-native

Christoph Kappel · · , , , , , , , · Reading time 7 minutes

I was never a big fan of all these cloud stuff, especially not of the cloud-native idea, but for my current project I have to get over and used to it.

So far, my only really experience with Kubernetes was during universities, guided training sessions at work, with some experienced instructor. That never worked well for me.

I can do guided stuff all day long, but when I don’t get my hands dirty, this stuff never sticks to me. Most of the commands succeed and I have no idea what I succeeded in and/or what is supposed to work now.

Well - I can surely read the next chapter of the instructions, but my learning just don’t works this way.

Initial thoughts &

For me, the whole stack is a bit weird and I start with some questions?

  • Where do I start?

  • Do I need some real installation of Kubernetes?

  • What about the bundled one from "Docker for Desktop"?

  • Why do I have to use macOS for this kind of stuff?

After some asking and searching, I made some decisions: I want to stick with the k8s docker brings to the party, along with Kind, for the ease of the management of clusters. I might delve into the Helm odyssey later.

And I going to use Quarkus as a passenger, there are some extensions, that might make the journey worthwhile.

I’ll skip all the install stuff of Docker and Kind, I think there is more than enough available and focus on the fun stuff and expect everything is ready.

Quarkus &

Scaffolding a Quarkus project is fairly easy:

$ mvn io.quarkus:quarkus-maven-plugin:1.9.1.Final:create \
-DprojectGroupId=dev.unexist.example \
-DprojectArtifactId=quarkus-hello-example \
-DprojectVersion=0.1 \
-DclassName="dev.unexist.showcase.HelloResource" \
-Dextensions="health, quarkus-smallrye-openapi, container-image-jib, kubernetes"

After that, you end up with a hello project and some helpful extensions:

  • health provides the required health and readyness for the k8s pod

  • openapi generates the OpenAPI3 and bundles the Swagger-Ui

  • container-image-jib actually builds the images without a requirement for actual Docker

  • kubernetes creates the helpful k8s manifests

Additional config &

Still, a bit config is required to create the ingress manifests, have a proper path and to really include the swagger-ui in everything instead of dev builds only. So without further ado, can you please add the three lines to your application.properties file?

quarkus.kubernetes.expose=true
quarkus.servlet.context-path=/hello
quarkus.swagger-ui.always-include=true

Container time &

And following builds the container, pushes it to the local docker and generates the some helpful k8s manifests in one command:

$ mvn clean package -Dquarkus.container-image.build=true

Manifests &

The kubernetes extension created nice manifests for us, which can be found here:

quarkus-hello-example/target/kubernetes

Kind &

After the passenger is ready, we have to set up our cluster. During my experiments I had to do this several times, because I only lated realized I forgot something. So I daresay we’ll do it right on the first time and init our cluster directly with the necessary stuff like ingress:

Create a cluster & 

cat <<EOF | kind create cluster --name example --config=-
kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
  kubeadmConfigPatches:
  - |
    kind: InitConfiguration
    nodeRegistration:
      kubeletExtraArgs:
        node-labels: "ingress-ready=true"
  extraPortMappings:
  - containerPort: 80
    hostPort: 80
    protocol: TCP
  - containerPort: 443
    hostPort: 443
    protocol: TCP
EOF

This command creates a new cluster named example and sets up some magic port mapping required for ingress.

Docker images &

In order for Kubernetes to find our image, we have to load it first. That can be done like this:

$ kind load docker-image docker.io/unexist/quarkus-hello-example:0.1 --name example

Kubernetes &

Docker and kind have done their best to make it really easy for us. So let’s go on.

Dashboard &

I have no problems with the CLI of k8s or tools like k9s, but a nice dashboard with some fancy graphs and a way to see all at once is a quite nice.

Installation &

The current version at writing this is 2.0 and can be installed with the next line:

$ kubectl apply -f https://raw.githubusercontent.com/kubernetes/dashboard/v2.0.0/aio/deploy/recommended.yaml

User account &

Once the installation is done we need some accounts to access our new dashboard, the next two manifests create an admin for it:

cat <<EOF | kubectl apply -f -
apiVersion: v1
kind: ServiceAccount
metadata:
  name: admin-user
  namespace: kubernetes-dashboard
EOF

And..

cat <<EOF | kubectl apply -f -
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: admin-user
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: cluster-admin
subjects:
- kind: ServiceAccount
  name: admin-user
  namespace: kubernetes-dashboard
EOF

Run it &

In order to access the dashboard, a running proxy is required:

$ kubectl proxy

Log in - finally! &

The easiest way to log into this dashboard is via a token, this can be fetched via CLI like this:

$ kubectl -n kubernetes-dashboard describe secret $(kubectl -n kubernetes-dashboard get secret | grep admin-user | awk '{print $1}')

Copy this token and use it here:

http://localhost:8001/api/v1/namespaces/kubernetes-dashboard/services/https:kubernetes-dashboard:/proxy/#/login

Ingress &

We created the cluster with support for ingress, but so complete the installation another quick step is required.

Finishing up &

Last step: Init our ingress controller:

$ kubectl wait --namespace ingress-nginx \
  --for=condition=ready pod \
  --selector=app.kubernetes.io/component=controller \
  --timeout=90s

To be continued.