API SPECIFICATION

The application endpoints:

• /

  • Accepts GET requests.

  • Returns a json body response

  • Example use:

    curl 127.0.0.1:8080

  • Example response:

    {"message": "Hello Smile"}

• /healthy

  • Accepts a GET request

  • Returns 200 if the application is ready.

  • Returns 500 if the application is not ready

  • Example use:

    curl 127.0.0.1:8080/healthy

INFRASTRUCTURE

The App is packaged as a docker image and can be deployed on kubernetes. The basic manifests required to deploy the application to a kubernetes cluster are provided in the k8s folder

    └── k8s   
        ├── deployment.yaml
        ├── svc.yaml
        ├── hpa.yaml
        ├── namespace.yaml
        └── sa.yaml

Kubernetes Resources deployed

1. ServiceAccount
2. Deployment
3. Service
4. HorizontalPodAutoscaler
5. Namespace

AWS Resources deployed

1. VPC|Subnets|NatGateways
2. EKS Cluster
3. ECR Repository
4. IAM Resources

Production

Deployment Steps

NOTE: Terraform state is stored locally !!

Requirements

• Docker >= 1.24 with Compose
• Terraform >= 1.2.3
• AWS cli >= 2.13.11
• yq >= 4.40.5
• kubectl >= 1.28.2

Tip: Use the helper script to run all deployment steps with 1 command. You must first set environment variables in step 1.

By default the script will use IMAGE_TAG=1.0.0 , to use a different tag run export IMAGE_TAG=<preferred tag> first.

 ./helper.sh k8s-deploy

Individual Steps

1. Set up AWS credentials. You can read how here

   Recommended that you use the option of setting environment variables for this demo

   export AWS_SECRET_ACCESS_KEY=<Actual Secret Access Key>
   export AWS_ACCESS_KEY_ID=<Actual Access Key> 
   export TF_VAR_aws_region=<Actual AWS Region> 

2. Change directory to the terraform folder

   cd terraform

3. Initialize Terraform configuration by using the command below

   terraform init 

4. Plan Terraform configuration by using the command below

   terraform plan

5. Apply Terraform configuration by using the command below

   terraform apply 

   Confirm the planned changes by entering yes

   Do you want to perform these actions?
   Terraform will perform the actions described above.
   Only 'yes' will be accepted to approve.
   
   Enter a value: yes

6. Set the IMAGE environment variable

   export IMAGE_REPO=$(terraform output  -raw container_repository)
   export IMAGE_REGISTRY=$(echo ${IMAGE_REPO}| cut -d '/' -f1)
   export IMAGE_TAG=1.0.0
   export IMAGE=$IMAGE_REPO:$IMAGE_TAG

7. Build and Push image to container registry

   # change directory
   cd ../
   
   # build image
   docker build -t simple-app:$IMAGE_TAG .
   
   # tag image
   docker tag simple-app:$IMAGE_TAG $IMAGE
   
   # login to image registry 
   aws ecr get-login-password --region $TF_VAR_aws_region | docker login --username AWS --password-stdin $IMAGE_REGISTRY
   
   
   # push image to registry
   docker push $IMAGE
   

8. Deploy kubernetes manifests

   # set image and tag in mainifest
   yq e -i '.spec.template.spec.containers[0].image=env(IMAGE)' k8s/deployment.yaml
   
   # set kube context
   aws eks update-kubeconfig --region $TF_VAR_aws_region --name main --alias main 
   
   # apply namespace manifests
   kubectl apply -f k8s/namespace.yaml     
   
   # apply other manifests
   kubectl apply -f k8s/   
   

Cleanup Steps

Tip: Use helper script to run all cleanup steps with 1 command.

 ./helper.sh k8s-clean

Individual Steps

1. Follow the steps below to cleanup the created resources

   # delete kubernetes resources
   kubectl delete -f k8s/ 
   
   # Force delete ECR Repository using aws cli. (Terraform cannot force delete because the repository is not empty)
   aws ecr delete-repository --repository-name simple-app --force --region $TF_VAR_aws_region
   
   # Destroy terraform resources
   cd terraform
   
   terraform destroy

2. Confirm that you want to delete the resources

   Do you really want to destroy all resources?
   Terraform will destroy all your managed infrastructure, as shown above.
   There is no undo. Only 'yes' will be accepted to confirm.
   
   Enter a value: yes

Local Deployments

1. Deploying as a local docker container

Requirements

• Docker with Compose
• Port (8080)

To deploy as a local docker container you can use the helper script provided by running the below command in the project root directory

 ./helper.sh run

The script will do the following:

1. Builds the docker image with tag simple-app:latest
2. Runs a local container with the image
3. Exposes the container for local network traffic on port 8080

Access the application on http://localhost:8080

You can clean up the local docker container by running

./helper.sh clean

1. Running locally in debug mode

Requirements

• python3
• python3-pip
• Port (8080)

To run the app locally in debug mode you can use the helper script provided by running the below command in the project root directory

 ./helper.sh debug

The script will do the following:

1. Runs the python app locally on port 8080

Access the application on http://localhost:8080

Road Map

Some features that should be implemented in the future

• Add Node Autoscaling Capability to the cluster (Karpenter, Cluster Autoscaler)
• Implement a Service Mesh
• Observability Dashboard for monitoring resources in the cluster
• Policy As Code Tooling (OPA, Kyverno)
• Implement Topology Spread Strategy for Nodes across Multiple AZs
• Deploy an Ingress Controller (Possible use Mesh Gateway)