Boilerplate for your next web project.

Rocketplate for your next adventure 🚀

To get started, create a Google Cloud project and get authenticated:

gcloud auth application-default login
gcloud auth configure-docker

Now, we need to wire up Terraform and GCS. You can use the ./bootstrap.bash tool to do this:

./bootstrap.bash TERRAFORM_BUCKET_NAME

which will create a GCS bucket for Terraform and authenticate Google Cloud Build.

To configure your new deployment, copy the example tfvars:

cp -n example.tfvars production.tfvars

and change anything that needs to be changed.

All that's left is to deploy!

# Replace TERRAFORM_BUCKET_NAME
terraform init --backend-config=bucket=TERRAFORM_BUCKET_NAME --backend-config=prefix=tfstate/production

# You might need to run this twice
terraform apply -var-file=production.tfvars

You now are admin-ing your own GKE cluster 😎

To get your domain name pointing at this cluster, point the domain name's nameservers at the Google Managed Zone's DNS servers.

You can get these by going to the Google Cloud UI > Network Services > Cloud DNS > [Your Managed Zone] > Registrar Setup (top right corner).

The tfvars files are .gitignored, so, to persist them, you can use the ./tfvars.bash tool:

STAGE=production TF_BUCKET=your-bucket-name ./tfvars.bash upload

Which will upload production.tfvars for you to your terraform bucket.

To have your team download, they can run:

STAGE=production TF_BUCKET=your-bucket-name ./tfvars.bash download

The CI system, Google Cloud Build, can now also download your secrets and deploy for you.

To use Google Cloud Build, first you need to build the custom cloud builder:

cd cloud-builder
gcloud builds submit --config=cloudbuild.yaml

Then, to run a remote deployment:

STAGE=production TF_BUCKET=rocketplate-terraform ./deploy.bash remote

To get builds happening automatically, head to the Google Cloud Build dashboard and set up an automatic trigger.

The database in this project is Postgres, and the migration system is rambler.

Rambler has it's own folder, in ./rambler/

To get the development environment up and running, run

docker-compose up

which will get Postgres and rambler running.

Next would be to get api and web and running:

cd api
cp -n .env.example .env
npm run install
npm run dev

cd web
cp -n .env.example .env
npm run install
npm run dev

The way it's set up, environment variables are read from .env in both services.

You can open up both localhost:3000 (web) and localhost:3001 (api) and start building stuff!

This system uses (from metal to user):

• Google Cloud Platform
• Terraform
• Postgres
• Kubernetes
• Helm
• LogDNA
• Prometheus
• Grafana
• rambler
• Node.js
• Micro
• React
• Next.js
• Kubernetes Nginx Ingress

To destroy everything, we can mostly just use Terraform. Doing a simple terraform destory fails unfortunatley though, because the Kubernetes node pool gets destroyed before helm (if you can create a PR to fix this, hats off to you!).

So, to destroy everything completely:

terraform destroy -var-file=production.tfvars
terraform state rm \
  module.ingress.helm_release.kube_lego \
  module.ingress.helm_release.nginx_ingress
terraform destroy -var-file=production.tfvars

And you should be good to go.

Try running this command.

• Move rambler deploy to terraform job resource when it lands: https://github.com/terraform-providers/terraform-provider-kubernetes/milestone/8

Whatever you do, work at it with all your heart, as working for the Lord, not for human masters, since you know that you will receive an inheritance from the Lord as a reward.

Colossians 3:23