Example of PKS based application deployment using Concourse

Base application to demonstrate PKS and Concourse pipelines

Functionality:

• In order to create Docker images, the pipeline is assuming that you have a Dockerfile in your repo.
• The kubernetes apply is assuming you have a deployment.yaml in your repo.

Building the docker image locally

• You can repurpose Dockerfile to fit your needs in terms of base image location and/or different set of scripts
• git clone this repo and from the cloned repo folder, execute:

git clone https://github.com/sharadg/hello-world-nodejs
cd hello-world-nodejs
docker build -t hello-world-nodejs:latest -f Dockerfile .

• Optionally, you can tag the docker image and push it to docker hub or Harbor registry

docker tag hello-world-nodejs:latest <harbor_repo_location>/library/hello-world-nodejs:latest
docker push <harbor_repo_location>/library/hello-world-nodejs:latest
# OR
docker tag hello-world-nodejs:latest <docker_hub_username>/hello-world-nodejs:latest
docker push <docker_hub_username>/hello-world-nodejs:latest

• Before deploying on k8s, confirm the location of image specified in deployment.yaml to make sure your k8s environment can access the harbor repo or docker hub as the case may be:

    spec:
      containers:
        - name: nodejs-app
          image: shagupta/hello-world-nodejs:latest
          imagePullPolicy: Always

• Using kubectl, you can then deploy the app on PKS or on locally running minikube:

kubectl apply -f deployment.yaml
kubectl get all

# If you are using minikube, you can find out the exposed service location
minikube service list 
# OR
minikube service nodejs-app-np --url

Using the example pipeline

• Create/populate a variables.yml file in your ci folder that holds credentials/locations for Docker hub or Harbor registry, and credentials for pks login and pks get-credentials
• fly the pipeline using

fly -t <concourse-target> set-pipeline -p <pipeline_name> -c ci/pipeline.yml -l ci/variables.yml

Credits

• I learned a lot from this blog post on deployment pipeline with kubernetes and concourse
• As well as the kubernetes-resource for concourse and the sample usage