Run Airy Core on AWS with Terraform

Requirements#

Terraform v1.0.0+
Airy CLI v0.34.0+
SSH key in ~/.ssh/id_rsa.pub
Kubectl (optional)

Create the Kubernetes cluster#

In case you already have a cluster running you can skip to the next section.

Amazon Web Services#

You need to provide Terraform with the AWS credentials for your IAM Role. If you don't know how to create one, follow these instructions. They have to be put into a terraform.tfvars file in terraform/kubernetes that looks contains:

aws_access_key = "<YOUR_AWS_ACCESS_KEY>"
aws_secret_key = "<YOUR_AWS_SECRET_KEY>"

If you want to deploy the EKS cluster in your existing VPC you have to override the vpc_id, subnets and fargate_subnets variables with the outputs from your vpc and set create_vpc=false in the Airy Core module.

terraform init
terraform apply

Install Airy Helm chart#

After the Kubernetes cluster has been created we can deploy the Airy Helm chart to it. Change to the terraform/main directory and make sure the .kubeconfig file is there.

You can configure your instance by making changes to infrastructure/terraform/main/files/values.yaml

If you want to deploy the stateless apps with AWS Fargate you can add workerType: fargate to the values.yaml

Finally, you need to initialize the Terraform workspace and run apply.

terraform init
terraform apply