What is this?

no more YAML. use your favorite programming language (JS, python and more) to build :

Build out the Fargate infrastructure for your  spring boot application, 
RDS,
including logging, 
Secret env variable in task definition (RDS Password),
cloudwatch dashboard and 
custom metric autoscaling with target tracking.

Prepwork

Use AWS Cloud 9 or Amazon workspace (need to add 8.8.8.8 in /etc/resolv.conf).

Create a DB password for RDS .. replace it with your own password value

aws ssm put-parameter --name "/mysqlpassword" --value "PassW0rd~~866" --type "SecureString"

Deploy using aws cdk

cd springboot-fargate-cdk
npm install

# do this only one time per region, this will bootstrap S3 bucket for CDK
npx [email protected] bootstrap 

#this will use cdk to deploy base infra, rds, and spring boot app on Fargate
npx [email protected] deploy  --require-approval never  "*"

Test the app:

This is a spring boot CRUD app with RDS (mysql).

# If you use workspace, you can install Postman.
# Otherwise, install httpie: https://httpie.org/doc/1.0.2
# on Cloud9: sudo pip install --upgrade httpie
# use httpie to POST to ALB endpoint / api/notes , eg:

http POST sprin-sprin-????????????.ap-southeast-1.elb.amazonaws.com/api/notes title="title 2" content="sample content"

# view the list of notes :

http sprin-sprin-????????????.ap-southeast-1.elb.amazonaws.com/api/notes

      Connection: keep-alive
      Content-Type: application/json;charset=UTF-8
      Date: Mon, 08 Jul 2019 06:15:10 GMT
      Transfer-Encoding: chunked

      [
          {
              "content": "aa",
              "createdAt": "2019-07-08T06:10:41.000+0000",
              "id": 1,
              "title": "title1",
              "updatedAt": "2019-07-08T06:10:41.000+0000"
          },
          {
              "content": "sample content",
              "createdAt": "2019-07-08T06:14:28.000+0000",
              "id": 2,
              "title": "title 2",
              "updatedAt": "2019-07-08T06:14:28.000+0000"
          }
      ]

Cloudwatch Dashboard

Pretty graph on Spring boot task cpu/memory, req/s, response time and a custom metric..

Cloudwatch log insight sample query - select the ECS task log group (tips: use Cloudformation console to find the task log group name)

fields @timestamp, @message
|filter @message like /JVM running for/
| sort @timestamp desc

Custom metric autoscaling

Trigger custom metric autoscaling:

while true; do aws cloudwatch put-metric-data --metric-name CDKTestingCustomMetric --namespace "CDK/Testing" --value $(( ( RANDOM % 10 ) + 180 )); sleep 60; done

Spring boot tasks should autoscale from 2 to 3,4,5...progressively. Watch the task count using Cloudwatch Dashboard. A max of 20 tasks will be launched as long as the custom metric stays above 150.

To scale it down run this command:

while true; do aws cloudwatch put-metric-data --metric-name CDKTestingCustomMetric --namespace "CDK/Testing" --value $(( ( RANDOM % 10 )  )); sleep 60; done

Clean up

npx [email protected] destroy