VAZ Projects Website

This is the repository for VAZ Projects, my personal blog and project log. Its main purpose is to serve as a sandbox for my DevOps learning journey.

Architecture Overview

The project consists of a Django application running on multiple EC2 instances within an Auto Scaling group. Traffic is routed to these instances using Traefik as the load balancer. A self-managed PostgreSQL instance is used as the database. Static and user-uploaded files are served through CloudFront.

GitLab is used for continuous integration and continuous delivery. Each new commit triggers a pipeline that builds container images, runs tests, and deploys the new version to a staging environment. Deployment to production can be started with manual approval.

A Grafana/Prometheus/Loki stack is used to monitor user activity, application performance and instance resource utilization, as well as ingesting logs for different services.

Running Locally

Install Dependencies

# Install Docker, Buildx, Docker Compose and mkcert.
pacman -S docker docker-buildx docker-compose mkcert	# On Arch Linux.

You might need to enable buildx by default with docker buildx install.

Project Setup

# Clone the repository.
git clone https://gitlab.com/marcelotsvaz/vaz-projects.git
cd vaz-projects/

# Create certificates with mkcert.
mkdir -p deployment/tls/

mkcert -ecdsa -install	# If not done yet.
mkcert -ecdsa -key-file deployment/tls/websiteKey.pem -cert-file deployment/tls/website.crt localhost minio

Start Compose Project

Make sure you have ports 80, 443, 8080, 9000 and 9001 available.

# Build the application image and start all containers.
docker compose up --detach --build

# Create account for Django admin.
docker compose run --rm application 'manage.py createsuperuser'

The following URLs will be available:

• Application: https://localhost/
• Django admin: https://localhost/admin/
• Traefik dashboard: http://localhost:8080/dashboard/
• MinIO console: https://localhost:9001/

Running Tests

# Run unit tests and generate coverage report.
docker compose run --rm --build application 'coverage run manage.py test && coverage report'

Development

# Mount the application folder and the compiled LESS file into the container so you can test changes without rebuilding the image.
docker compose -f compose.yaml -f development.compose.yaml up --detach --build

Cleanup

# Stop and remove all containers, data is preserved in named volumes.
docker compose down

# Stop and remove all containers, including volumes.
docker compose down --volumes