This project provides a web service that generates random names to identify objects among humans.

It provides two collections :

• fr : names and french adjectives
• docker : same names and adjectives as the Docker project

• AWS is used for hosting all components
• Terraform is used for provisioning infrastructure
• Name generator service is configured to run, as a docker container, in Elastic container service (ECS)
• AWS CodeDeploy is used for deployment management

Terraform is used for infrastructure provisioning. All infra related code resides in tf/
Terraform Cloud is used for IaC pipeline management.

Infrastructure components:

• ECS - hosting the app as a container
• Application Load Balancer (ALB) - used as a router and load balancer
• CodeDeploy - used for deployment management

Below is the simplified infra diagram:

For this purpose Github Actions are being used. Pipeline scripts are stored in .github/ directory.

Building and deploying of the application happens in a following way:

1. On every push to the master branch pipeline is being invoked
2. Name generator application is built stored as a docker image
3. Built docker image is pushed to ECR
4. CodeDeploy helper lambda functions are being built and deployed
5. CodeDeploy Blue/Green deployment is created and invoked

CodeDeploy is performing Blue/Green deployment in a following way:
Step 1: Deploying replacement task set
Step 2: Test traffic route setup
After this step, Lambda function will validate the setup of test task set
Step 3: Rerouting production traffic to replacement task set
After this step, Lambda function will validate the setup of production task set
Step 4: Waiting for deployment to stabilize
Step 5: Terminate original task set

More info: https://docs.aws.amazon.com/codedeploy/latest/userguide/deployment-steps-ecs.html#deployment-steps-what-happens

Below is a simple deployment diagram in case of Blue/Green deployment strategy:

Before initiating the deployment check the following:

1. Config files used in deployment execution:
   a. .github/workflows/codedeploy/appspec.yaml - info about the service to be deployed and validation lambda functions
   b. .github/workflows/codedeploy/lambda - lambda functions to be used for validating test and production task sets
   c. .github/workflows/codedeploy/create-deployment.yaml - info for creating and initiating new deployment
2. .github/workflows/cd.yaml - deployment pipeline definition

1. Push changes to the master branch.
2. Monitor the build and deployment pipeline in Github Actions: https://github.com/nediml/name-generator/actions
3. Monitor the deployment execution in CodeDeploy: https://console.aws.amazon.com/codesuite/codedeploy/deployments?region=us-east-1

Note that it is possible to cancel and rollback at any point using CodeDeploy console at the link above.

Application is reachable at:

• Prod task set: https://tf-lb-2021110609264334060000000b-792189563.us-east-1.elb.amazonaws.com:80
• Test task set: https://tf-lb-2021110609264334060000000b-792189563.us-east-1.elb.amazonaws.com:81

1. Push changes to the master branch to te tf/ directory
2. Go to https://app.terraform.io/app/lokalise-interview/workspaces
   • view terraform plan results
   • set the values for terraform variables
3. Approve/Reject the plan accordingly