Gatekeeper

A production-style API gateway in Python/FastAPI with API key authentication, per-key rate limiting, and structured request logging.

Live URL: https://gatekeeper-ccgq.onrender.com/docs

Features

• API Key Authentication — Validates incoming requests via the X-API-Key header against a cloud-hosted datastore and returns 401 Unauthorized for invalid credentials.
• Per-Key Token-Bucket Rate Limiting — Enforces traffic limits using configurable token-bucket parameters stored in the database.
• Resilient Async Proxy — Forwards authenticated traffic to upstream services with timeout protection and proper error handling.
• Structured Request Logging — Captures request metadata including route, status code, latency, and client IP for observability and auditing.
• Environment Isolation — Supports Local, CI, and Production environments through environment-variable based configuration.

Architecture & Stack

• Framework: Python 3.12, FastAPI, Uvicorn
• ORM: SQLAlchemy 2.0
• Migrations: Alembic
• Database: Neon PostgreSQL
• Deployment: Docker, Render
• Testing: Pytest, GitHub Actions

Endpoints

Method	Endpoint	Description
GET	/	Health check
GET	/docs	Swagger UI documentation
GET	/proxy	Authenticated, rate-limited proxy endpoint

---

Infrastructure Deployment

[ Local Client / curl ] --- (HTTPS) ---> [ Render Container ]
                                                |
                                         (Auth & Logging)
                                                v
                                     [ Neon PostgreSQL ]

Production Configuration

• DATABASE_URL — PostgreSQL connection string
• PORT — Container port (default: 8000)

---

Local Development

1. Configure Environment Variables

Create a .env file:

DATABASE_URL="postgresql://postgres:postgres@localhost:5432/gatekeeper"
TEST_DATABASE_URL="postgresql://postgres:postgres@localhost:5432/gatekeeper_test"

2. Start PostgreSQL

docker-compose up -d

3. Apply Migrations

alembic upgrade head

4. Run the Application

uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

5. Run Tests

pytest