vihangpatil/http4k-k8s

Bootstrap http4k minimal server app in different stages of porting to k8s

http4k on k8s

Bootstrapped minimal http4k app with inbuilt support for k8s using this guide.

Different stages of transition of the app onto k8s

1. Run locally using gradle
2. Run using docker
3. Run using docker compose
4. Run on k8s using kubectl
5. Run on k8s using helm

For steps 1 to 3, you reach application by

curl -v http://localhost:8081/liveness
curl -v http://localhost:8081/readiness
curl http://localhost:8080/ping

For step 4 & 5, only http://localhost:8080/ping is exposed.
liveness and readiness probes are not exposed.
Instead they are used internally by k8s.

1. Run locally via gradle

./gradlew run

2. Run in Docker

2a. Build app for docker

./gradle install

2b. Build docker image

docker build --tag backend .

docker images

2c. Run in docker

Use --detach instead of --rm to run in background.

docker run \
  --name backend \
  --publish "8080:8080" \
  --publish "8081:8081" \
  --rm \
  backend

docker ps -a

docker rm -f backend

3. Running in docker compose

docker compose up --build

docker ps -a

docker compose down

4. Run on local k8s using kubectl

Enable k8s in Docker Desktop for macOS.

Check which k8s you are connected to.

kubectl config get-contexts
kubectl config use-contexts docker-desktop

kubectl get nodes

If you are not using [default] namespace, you will have to pass --namespace {{custom_namespace}}.

[Optional] Install k8s dashboard

kubectl apply -f https://raw.githubusercontent.com/kubernetes/dashboard/v2.7.0/aio/deploy/recommended.yaml

kubectl apply -f k8s/k8s-backend.yaml 

kubectl get services
kubectl get deployments
kubectl get pods

kubectl delete service backend-service
kubectl delete deployment backend-app

5. Run on local k8s using helm

helm install backend-app helm/backend