Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/DevOpsDuoc/Evaluacion02_Devop_Innovatech/llms.txt

Use this file to discover all available pages before exploring further.

Innovatech Chile runs across three EC2 instances in a private VPC. Deployments happen automatically on every push to main via the GitHub Actions pipeline, or manually using the shell scripts in proyect/remote-setup/. This page covers both paths and explains how to SSH into private instances through the public bastion.

​
EC2 architecture

InstanceTierNetworkServices
ec2-webWeb / BastionPublic subnet — Elastic IPtienda-frontend on port 80
ec2-appApplicationPrivate subnettienda-backend on port 3001, tienda-backend-despachos on port 3002
ec2-datosDataPrivate subnetMySQL database on port 3306
ec2-web is the only instance with a public-facing Elastic IP. Use it as a bastion to reach the private instances.
Terraform prints the IP addresses of all instances as outputs. Run terraform output after provisioning to get the Elastic IP for ec2-web and the private IPs for ec2-app and ec2-datos.

​
Automatic deployment

Pushing to main triggers the GitHub Actions workflow, which:
  1. Builds and pushes updated images to Amazon ECR.
  2. Sends SSM RunShellScript commands to ec2-web and ec2-app by their Name tag.
  3. Each instance pulls the latest image and restarts the container.
No SSH access or open inbound ports are required for automated deployments. See the pipeline page for the full workflow file and secret configuration.

​
Manual deployment

Use the manual path when you need to deploy without triggering a push, or when setting up an instance for the first time.

​
SSH access

1

Connect to ec2-web (bastion)

Use the Elastic IP printed by terraform output web_eip_public_ip.
ssh -i your-key.pem ec2-user@<ELASTIC_IP>
2

Jump to ec2-app or ec2-datos

From inside ec2-web, use the private IPs printed by terraform output.
# Jump to the application layer
ssh -i your-key.pem ec2-user@<APP_PRIVATE_IP>

# Jump to the data layer
ssh -i your-key.pem ec2-user@<DATOS_PRIVATE_IP>

​
SSM Session Manager (no SSH key required)

If you have the AWS CLI and the Session Manager plugin installed, you can open a shell without any SSH key or open inbound port: 
aws ssm start-session --target <INSTANCE_ID> --region us-east-1

​
First-time instance setup

Run 00-init.sh once on each instance to install Docker before deploying any containers. See deployment scripts reference for the full script content. 
bash proyect/remote-setup/00-init.sh
After the script finishes, log out and back in so the docker group membership takes effect.

​
Deploy containers manually

1

Authenticate with ECR

Set your AWS account ID and log Docker into the private registry.
export AWS_ACCOUNT_ID=118812498736
aws ecr get-login-password --region us-east-1 \
  | docker login --username AWS --password-stdin \
    ${AWS_ACCOUNT_ID}.dkr.ecr.us-east-1.amazonaws.com
2

Copy docker-compose.yml to the instance

Place the Compose file in /home/ec2-user/app/ — this is the working directory expected by 01-pull_and_deploy.sh.
scp -i your-key.pem proyect/docker-compose.yml \
  ec2-user@<ELASTIC_IP>:/home/ec2-user/app/docker-compose.yml
3

Set the AWS_ACCOUNT_ID environment variable

The Compose file and deploy script reference this variable to construct ECR image URLs.
export AWS_ACCOUNT_ID=118812498736
4

Pull and start all containers

Run 01-pull_and_deploy.sh from inside the instance. The script authenticates with ECR, pulls the latest images, and starts the full stack.
bash proyect/remote-setup/01-pull_and_deploy.sh
This runs docker compose pull followed by docker compose up -d --remove-orphans in /home/ec2-user/app/.
Make sure docker-compose.yml is present in /home/ec2-user/app/ before running 01-pull_and_deploy.sh. The script will exit with an error if that directory does not exist or the file is missing.

Build docs developers (and LLMs) love

Get started for freeTalk to us