Cloud deployment (AWS)

Microservice stack supports cloud deployment out of the box. When you start your Microservice stack project using Configuration generator generates infrastructure/cloud folder. This folder contains Terraform configuration required to deploy the whole stack to AWS with just few commands.

In order to deploy your application to the cloud, you will need to have Terraform (https://www.terraform.io/) installed and an AWS account setup with a domain parked inside Route53.

Once your AWS account is set up, proceed to install AWS CLI on your machine and configure it. You can follow the following guide to do that https://docs.aws.amazon.com/cli/latest/userguide/getting-started-prereqs.html.

After the AWS CLI is set up, you can proceed with installing the infrastructure. Move to the infrastructure/cloud folder and execute the following commands, which will set up the Terraform environment

export TF_VAR_access_key=<AWS ACCESS KEY ID>
export TF_VAR_secret_key=<AWS SECRET KEY ID>
export TF_VAR_region=<AWS REGION>

Once this is done, go into the variables.tf file and change variables to your configuration. Bellow is brief explanation of that these variables mean

• Environment name The name of your environment/deployment. Resources such as VPC and EKS will be named after this.

• Domain Top level domain of your application. When installing the configuration, api subdomain will be automatically added to the domain value, so you don't have to.

• Zone ID The ID of Route53 zone the domain is located in.

• Output secrets

  Determines whether secrets such as database credentials, and Github actions security keys should be output locally into JSON files.

• Load balancer URL

  Load balancer DNS record, this should stay empty on the first Terraform run. After the first run is complete, you will be able to copy the load balancer URL from AWS console and replace the variable. After that run the terraform apply again, and api subdomain will be pointed to the load balancer.

• Deploy RabbitMQ

  Determines whether to deploy RabbitMQ or not.

• Deploy Redis

  Determines whether to deploy Redis or not.

• Deploy PostgreSQL

  Determines whether to deploy PostgreSQL database or not.

First Terraform run

After variables are updated, you can run terraform apply command. This command will create a plan of what will be changed inside the AWS and upon confirmation, it will create all the shown resources inside the AWS.

After the Terraform job has finished, you can verify that your cluster has been setup by getting Kubeconfig using the following command

aws eks update-kubeconfig --name <cluster-name> --region <region>

Please keep in mind that the cluster name is the same as the environment_name variable you set inside the variables.tf.

When AWS CLI command completes, you can see the cluster using the kubectl cli.

kubectl get nodes

Second Terraform run

With the cluster setup, we can now finalize the deployment by providing the load_balancer_url variable to variables.tf file. To get the load balancer url go to EC2->Load Balancers and copy the newly created load balancers DNS name. With the variables updated, you can again run terraform apply command to apply the changes and finish AWS configuration.

Github CI

The cluster is now setup and accessible, the only thing left to do is set up the Github CI secrets, so you can start deploying your application.

If you opted for output_secrets option, all the secrets required for Github CI configuration will be located inside the infrastructure/cloud/outputs folder. If not, you will have to get the secrets by looking into the terraform state.

To setup the secrets go to your Github repository settings, and select Security->Secrets and variables->Actions. There you can add new repository secrets

• AWS_ACCESS_KEY_ID Github actions IAM users access key, located in outputs/iam_user_credentials.json

• AWS_SECRET_ACCESS_KEY Github actions IAM users secret key, located in outputs/iam_user_credentials.json

• AWS_REGION The AWS region you deployed to

• CLUSTER_NAME The name of the cluster, should be the same as environment_name variable

• AWS_ECR_REPOSITORY Should equal <CLUSTER_NAME>_container_repository

After these secrets are set, you can deploy your applications by tagging your code using v*.*.* tag.