Java AWS CDK Stack for a Spring Boot App that uses a Database

This AWS CDK Stack deploys a Spring Boot based app to ECS. It assumes the image is a Spring boot that connects to a database.

Code Repository

https://github.com/mlopezsoto/aws-cdk-volunt-ecs

Requirements

• AWS Account
• Java AWS CDK project.

To install AWS CDK follow the steps outlined here: https://cdkworkshop.com/

Stack Steps

• VPN creation with two subnets, one public one isolated (software.amazon.awscdk.services.ec2.Vpc).
• Postgres Database creation (software.amazon.awscdk.services.rds.DatabaseInstance).
• ECS Fargate Cluster creation (software.amazon.awscdk.services.ecs.patterns.ApplicationLoadBalancedFargateService).

Usage example

Once you have your AWS CDK project, you can copy the class AwsCdkSpringPostgresEcsFargateStack to your project and use it. For instance:

import software.amazon.awscdk.App;

public class AwsCdkSpringPostgresEcsFargateApp {
    public static void main(final String[] args) {
        App app = new App();
        
        AwsCdkSpringPostgresEcsFargateStack.Builder.create(app, "butterfly", "butterflydb", "butter", "flyingbutter", "'docker.io/mlopezsoto/volunt:latest'")
                .ecsTaskCpu(256)
                .ecsTaskMemoryLimitMiB(512)
                .ecsTaskDesiredCount(1)
                .springContainerPort(8080)
                .targetGroupHealthCheckPath("/actuator/health")
                .vpcMaxAzs(3)
                .build();

        app.synth();
    }
}

The code above creates the folowing:

• CDK/Cloudformation Stack named 'butterfly' (Parameter 2)
• Database named 'butterflydb' (Parameter 3)
• Database username 'butter' (Parameter 4)
• Database password 'flyingbutter' (Parameter 5)
• Pulls and runs image from 'docker.io/mlopezsoto/volunt:latest'

Using the values above as an example, you can deploy your stack using:

cdk deploy butterfly

After it finishes, it will display some results and your application will be available online. Your application should be available from the Load Balancer DNS name, which you can get from the output 'ALBDNSName'. The prefix changes depending on the name you use for your stack, in our case it is 'butterfly'.

Outputs:
butterfly.ALBDNSName = butte-butte-DZD07FFTQKI5-224749273.ap-southeast-2.elb.amazonaws.com
butterfly.DatabaseAddress = butterfly-database-postgresql.cmwkrmicwxfc.ap-southeast-2.rds.amazonaws.com
butterfly.DatabasePort = 5432
butterfly.butterflyecsserviceLoadBalancerDNS7FF68FBE = butte-butte-DZD07FFTQKI5-224749273.ap-southeast-2.elb.amazonaws.com
butterfly.butterflyecsserviceServiceURL1D691D62 = http://butte-butte-DZD07FFTQKI5-224749273.ap-southeast-2.elb.amazonaws.com
butterfly.datasourceUrl = jdbc:postgresql://butterfly-database-postgresql.cmwkrmicwxfc.ap-southeast-2.rds.amazonaws.com:5432/butterflydb

Sample App

The image 'docker.io/mlopezsoto/volunt:latest' contains a basic Spring Boot app that creates a user table and exposes a couple of endpoints:

• /user/listAll
• /user/validateCredentials?username=&password=

You can test this using the CDK output 'butterfly.ALBDNSName' (sample above) and then appending the above paths.

IF butterfly.ALBDNSName = butte-butte-DZD07FFTQKI5-224749273.ap-southeast-2.elb.amazonaws.com THEN: http://butte-butte-DZD07FFTQKI5-224749273.ap-southeast-2.elb.amazonaws.com/user/listAll (This link won't work. The stack used to generate this does not exist anymore)

App code: https://github.com/mlopezsoto/volunt-be

Note: You can use your own image. This stack will deploy your image and set environment vars for your database access.

Basic CDK commands

List all Stacks

cdk ls

Deploy a stack, for instance a stack named 'butterfly' as in the sample code above

cdk deploy butterfly

Destroy (undeploy) a stack, for instance a stack named 'butterfly' as in the sample code above

cdk destroy butterfly