Deploy to Kubernetes (Helm)

Useful Links¶

Deployment of a `sbt`-built app on Kubernetes (MiniKube)¶

Test packaging without Kubernetes first¶

Stage all Play files in a local directory and verify

sbt stage

For direct deployment, create a distribution in target/universal

sbt dist

The dist task builds a binary version of your application that you can deploy to a server without any dependency on SBT, the only thing the server needs is a Java installation.

Deploy a Helm chart to Kubernetes¶

Prerequisites: minikube, kubectl, docker client and helm should be installed

Generate the Dockerfile and environment prepared for creating a Docker image

sbt docker:stage

Verify the output under target/docker
Start minikube

minikube start

Enable Ingress

minikube addons list
minikube addons enable ingress

Also consider enabling heapster

List available nodes to verify that kubectl is properly configured

kubectl get nodes

It should return one node.

Connect the Docker client to the Docker daemon in the K8s VM

eval $(minikube docker-env)

Just make sure you tag your Docker image with something other than ‘latest’ and use that tag while you pull the image. Otherwise, if you do not specify version of your image, it will be assumed as :latest, with pull image policy of Always correspondingly, which may eventually result in ErrImagePull as you may not have any versions of your Docker image out there in the default docker registry (usually DockerHub) yet.

If needed, remove previously built images from the local Docker server with sbt docker:clean or docker rmi <image>. To view the list of Docker images, run docker images
Build the Docker image and publish it to Kubernetes' Docker server.

sbt docker:publishLocal

Deploy the Helm chart

./helm install --dry-run --debug <helm chart folder> &> output.txt

and if that looks OK

./helm install <helm chart folder>

or specify a release name:

./helm install --name <release name> <helm chart folder>

Verify the Helm deployment to minikube

./helm list
./helm status <release name>

More details via:

kubectl get ing
kubectl get service
kubectl get deployment
kubectl get pods

Test the deployment by forwarding a local port to a port on the pod

kubectl get pods
kubectl port-forward <pod name> 8080:<target port on pod> 
curl -v http://localhost:8080/api

kubectl port-forward also allows using resource name, such as a service name, to select a matching pod to port forward to

kubectl port-forward svc/<service name>  8080:<service port> 
curl -v http://localhost:8080/

When needed, delete the release with

helm ls
helm delete <release name>

If you want to deploy / not deploy an Ingress¶

Update values.yaml in the Helm chart root folder

ingress:
  enabled: true  # or: false; true by default

If true, make sure the minikube Ingress add-on is enabled

minikube addons enable ingress

Deploy on Kubernetes as above

See Blog

SSL Termination (TO DO)¶

IBM Ingress TLS tutorial

Generate a x509, pem encoded, RSA 2048 certificate with OpenSSL

openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout ${KEY_FILE} -out ${CERT_FILE} -subj "/CN=${HOST}/O=${HOST}"

openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout tls.key -out tls.crt -subj "/CN=john-cd.com"

Note: To find myhost.com for minikube, run the following commands:

 $ minikube ssh
 $ echo $HOSTNAME
 minikube

Create a Kubernetes secret

kubectl create secret tls ${CERT_NAME} --key ${KEY_FILE} --cert ${CERT_FILE}

kubectl create secret tls my-secret --key tls.key --cert tls.crt

Add under spec: in

  tls:
  - hosts:
    - myhost.com
    secretName: my-secret

Find and delete all nginx pods to force the nginx.conf to update and reflect the ingress changes. Find the ingress pods with the following:

kubectl get pods --all-namespaces
kubectl delete pods --namespace=kube-system [ingress pod]