kubernetes-workshop

This repository is meant to be the starting point for students who are attending the kubernetes workshop at 42 wolfsburg.

Prereqisites:

• docker installed
• golang installed
• kubectl installed

Create a local kubernetes cluster with kind

• run go install sigs.k8s.io/kind@v0.23.0 to download kind
• run kind create cluster --name first-k8s-cluster to create a local kubernetes cluster in a docker container with kind
• run docker ps to see the created control plane container which is being used to interact with the kubernetes API
• run kind get nodes or kubectl get nodes to see that the above container represents a node in your local kubernetes cluster
• for now, you can delete your cluster with the command kind delete cluster --name first-k8s-cluster

Deploy a NGINX webserver on a configured cluster

• when creating a kind cluster, we can pass a config file like kind-cluster.yaml in order to add certain configurations for e.g. accessing the cluster right from the start.
• create a fresh cluster with kind create cluster --name second-k8s-cluster --config kind-cluster.yaml
• have a look at the webserver.yaml file, this is the webserver we will create and expose on port 80
• run kubectl apply -f webserver.yaml to deploy the webserver to your cluster
• run kubectl get deployments to see that the deployment was successful
• run kubectl get pods to see that the application us up and running in the pod
• run kubectl get services to see that the service which will route traffic to our pod later
• run kubectl exec -it <POD_NAME> -- bash to log into that container
• inside the container, run curl localhost:80 to see that the nginx server is serving traffic as expected
• run exit to leave the container

Expose your webserver to the outside world (your computer)

• you can try to access your webserver from your laptops browser, but this will fail, as you didn't tell kubernetes to expose anything to outside of your cluster yet. It's similar to exposing ports of a docker container.

• we need something like a driver for our cluster through which it is technically enabled to receive traffic from the outside world

• we choose the popular NGINX ingress controller

• next, we create an ingress which will manage external access to the services in our cluster

• run kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/main/deploy/static/provider/kind/deploy.yaml to create the ingress controller (fyi: on GCP/AWS/Azure this is not needed, as ingress controllers are handled by the cloud provider)

• run k get all -n ingress-nginx to see all the created resources (NOTE: the -n flag indicated that we look in the ingress-nginx namespace. Have a look at the concept of namespaces in kubernetes here)

• lastly, we define ingress rules for our cluster, such that our ingress controller knows, which traffic is allowed to enter the cluster. these rules are defined in ingress.yaml

• run kubectl apply -f ingress.yaml to apply the rules to your cluster (NOTE: the previous step takes a minute or two, just wait a bit and try again in case you get an error)

• run kubectl get ingress to see the newly created ingress and that it exposes port 80 of the cluster to the outside world!

• now your webserver is successfully exposed to the outside world of your cluster!

• run curl localhost:80 to see that you can access the server from your laptop. try the same address from your browser!

Scale up your NGINX instance

• with kubernetes we can easily make changes to our cluster and e.g. scale our replicas (amount of instances of a server in our case) up and down. This can be done via kubectl (kubectl scale deployment <deployment-name> --replicas=3), where the deployment name is nginx in our case. A more git-ops like approach and best practice is to put this configuration into our yaml files and check them into git. Change the replicas field in our webserver.yaml file to 2 and run kubectl apply -f webserver.yaml. This way you make use of the declarative approach of kubernetes!

Cleanup

• once you are done, you can delete your cluster with the command kind delete cluster --name second-k8s-cluster