MiniTwit (Chirp!) 🐦

A Twitter-like social media platform built with ASP.NET 9.0, deployed on DigitalOcean with full DevOps automation.

MiniTwit (nicknamed Chirp!) is a Group B exam project for the DevOps, Software Evolution and Software Maintenance BSc course at IT University of Copenhagen. It is a microblogging application where users can post short messages ("cheeps"), follow other users, and earn achievements — all backed by a production-grade DevOps pipeline.

🌐 Live application: https://minitwit.tech
📊 Monitoring (Grafana): http://209.38.113.150:3000/dashboards

Getting Started

Prerequisites

For local development:

• .NET 9 SDK
• Docker Engine or Docker Desktop (for the Compose option)

For infrastructure automation:

• Ansible
• Terraform / OpenTofu
• A DigitalOcean account + API token

---

Option 1: Run locally with .NET

The simplest path — uses SQLite automatically when no PostgreSQL connection string is set.

cd itu-minitwit
dotnet restore
dotnet build
dotnet run --project src/MiniTwit.Web

App available at: http://localhost:5273

---

Option 2: Run with Docker Compose

Starts the full stack — application, PostgreSQL, Prometheus, Grafana, Loki, and Alloy.

cp .env.example .env
# Edit .env and fill in POSTGRES_CONNECTION_STRING, GRAFANA_ADMIN_USER, GRAFANA_ADMIN_PASSWORD
docker compose up --build

Service	URL
Application	http://localhost:8080
Grafana	http://localhost:3000
Prometheus	http://localhost:9090
Loki	http://localhost:3100

---

Option 3: Provision infrastructure with Terraform

Provisions the full DigitalOcean infrastructure (droplets, firewalls, floating IP).

cd terraform
terraform init
cp terraform.tfvars.example terraform.tfvars
# Fill in terraform.tfvars with your DO token, SSH key, domain, etc.
terraform plan
terraform apply

---

Running Tests

The test suite includes unit, integration, and end-to-end (Playwright) tests.

cd itu-minitwit

# Restore and build first
dotnet restore
dotnet build

# Install Playwright browsers (needed for E2E tests)
pwsh tests/End2end/bin/Debug/net9.0/playwright.ps1 install chromium

# Run all tests
dotnet test

All tests must pass before any deployment — a failing test blocks the CI/CD pipeline.

---

Deploying to Production

Production deployments are fully automated via GitHub Actions + Ansible. A push to main triggers the full pipeline. You can also deploy manually:

Prerequisites

1. Servers provisioned via Terraform (see above)
2. GitHub Secrets configured (see below)
3. SSH access to the DigitalOcean droplets

Infrastructure Provisioning (Terraform)

If starting from scratch:

cd terraform
terraform init
terraform apply

Terraform will output the droplet IPs and auto-generate ansible/inventory.ini.

Application Deployment (Ansible)

To deploy manually without GitHub Actions:

cd ansible
ansible-playbook playbook.yml \
  -i inventory.ini \
  --extra-vars "image_tag=latest \
    dockerhub_username=<your_dockerhub_user> \
    dockerhub_token=<your_token> \
    postgres_connection_string='Host=<db_ip>;Database=minitwit;Username=minitwit_user;Password=<pw>;Port=5432' \
    grafana_admin_password=<password> \
    do_api_token=<token> \
    keepalived_auth_pass=<vrrp_password> \
    keepalived_virtual_ip=<floating_ip>"

Required GitHub Secrets

Configure these in Settings → Secrets and variables → Actions in your GitHub repository:

Secret	Description
SSH_HOST	IP of the primary LB droplet
SSH_HOST_SECONDARY	IP of the secondary LB droplet
SSH_USER	SSH user (typically root)
SSH_KEY	Private SSH key for server access
DOCKERHUB_USERNAME	Docker Hub account ID
DOCKERHUB_TOKEN	Docker Hub Personal Access Token
POSTGRES_CONNECTION_STRING	Full PostgreSQL connection string
GRAFANA_ADMIN_USER	Grafana admin username
GRAFANA_ADMIN_PASSWORD	Grafana admin password
LETSENCRYPT_EMAIL	Email for Let's Encrypt certificate registration
DO_API_TOKEN	DigitalOcean API token (for floating IP failover)
KEEPALIVED_AUTH_PASS	Keepalived VRRP authentication password
FLOATING_IP	DigitalOcean floating IP for HA failover

---

Contributing

We follow a trunk-based development model — all changes go to main via pull requests.

1. Fork the repository and create a feature branch:

   git checkout -b feature/my-feature

2. Make your changes. Ensure code is formatted:

   dotnet format

3. Run the full test suite locally:

   dotnet test

4. Open a pull request against main. The CI pipeline runs automatically on PR updates.
5. Once approved and merged to main, the CD pipeline deploys to production automatically.

Note: The pipeline will reject any PR that fails formatting checks, static analysis, or tests. Fix these locally before pushing.

---

Monitoring & Observability

The full observability stack is deployed alongside the application:

Tool	Purpose	URL (production)
Grafana	Dashboards & alerting	http://minitwit.tech:3000
Prometheus	Metrics collection	http://minitwit.tech:9090
Loki	Log aggregation	(via Grafana Explore)
Alloy	Log collection from Docker	internal
Node Exporter	System metrics (CPU, memory, disk)	internal

Grafana dashboards are provisioned as code from monitoring/grafana/dashboards/

---

Video Demonstrations - Made by mikcl

1. Infrastructure as Code (IaC)

---

2. CI/CD Pipeline

---

3. Monitoring

---

4. Centralized Logging