strapi-with-ecs

HERE IS YOUR WORKING URL FOR PROD AND STAGE

FOR PROD https://strapi.axceskart.com/admin/

FOR STAGE https://staging-strapi.axceskart.com/admin/

VERY-IMPORTANT-POINT AS OF NOW I DELETED ALL OF THESE RESOURCES BECUASE IT IS COSTING ME MORE THAN 300$. AND IT IS RUNING MORE THAN 2 WEEKS

IMPORTANT All requirements are fulfilled, all the sensitive informations are coming form secrets manager, as per best practise ECS cluster is in the private subnets , RDS cluster is also in provate subnets , applications are accecible only from Loadbalancer which is assoitaed with my domain.

STEPS TO CREATE THIS WHOLE PROJECT

NOTE HERE I AM USING MY OWN EC2 SERVER TO PERFORM THIS TASK (for example creating the apps, build the apps and push the apps...)

Step-1: First of all you need to create strapi project with customs settings and choose postgres-client, below you can find commands & screen shots...

npx create-strapi-app@latest strapi-with-pg

Step-2: After that go to this folder and create Dockerfile for containerize this strapi applications and create database.js file and .dockerignore file in here the content of database.js and .dockerignore file.

for database.js

module.exports = ({ env }) => ({
    connection: {
      client: 'postgres',
      connection: {
        host: env("DATABASE_HOST"),
        port: env.int("DATABASE_PORT"),
        database: env("DATABASE_NAME"),
        user: env("DATABASE_USERNAME"),
        password: env("DATABASE_PASSWORD"),
        ssl: {
          rejectUnauthorized: env.bool('DATABASE_SSL_SELF', false), // For self-signed certificates
        },
      },
      debug: false,
    },
  });

for .dockerignore

node_modules/
.tmp/
.cache/
.git/
build/

Step-3: Now build the dockerfile for both prod and stage and push into ECR repository...

Here is the command for build

docker build -t mystrapiopss:latest  --build-arg NODE_ENV=production .

Here ECR gives me how to push this image on ECR so i will not show here due to personal data informations...

NOTE one more thing is during the build it gives me lots of dependencies error, but after debugging and researching i found that the problem is in the node alpine base image so i changed from to official node:18.17.1-buster.

Step-4: So i already have me own domain which is hosted on Go-daddy, but according to your use-cases you want Route53 so i created and migrated my DNS nameserver from GO-daddy to Route53 Manually because DNS are very critical things...

Step-5: After that i created followinf resources...

1. S3 buckets (3 buckets) one for vpc , second is for prod and 3rd is for stage env for storing our terraform state files as per the best practise..
2. ACM wild-card certificates for our domain.
3. and 2 secrets in Secrets manger for storing sensitive infomation just like DB username , password, token etc. here you can find the screenshots.. these thing

Step-6: After that i go to networks folder which is to creating the VPC and use these terraform commands:-

terraform init

terraform plan

terraform apply

So in here our vpc is created with best practises...

Step-7: Then i go to prod folder and apply the same terraform commands inside aurora-db folder to create aurora-db postgress cluster with Multi-AZ & 2 instance one for writer and another is for reader.

Step-8: Then i go to ecs-cluster folder and apply the same terraform commands to create the ecs cluster with autoscaling

Step-9: Then same things for stage folder but in here difference is Aurora-db does not have Multi-AZ because this is the stage env, we dont want too much of cost.. and only 1 db-instance for writer & in ecs-cluster here i am using FARGATE_SPOT to save the cost.

Final-steps binded the prod and staging ALB into route53 with custom domain. see this..

Here you can see the applications logs, monitoring, and alarams, so through terraform i created a sns topic too for sending alerts to my email