Rate Limiter Service

A lightweight rate limiting service that supports multiple rate limits for different keys and tracks usage statistics.

Features

• Multiple Rate Limits: Configure different rate limits for different keys
• Statistics Tracking: Track hourly and daily usage for each key
• In-Memory Storage: Simple in-memory storage for both rate limits and statistics
• Health Checks: Endpoint to verify service health

API Endpoints

Check Rate Limit

GET /check/{key}

Checks if a key is within its rate limit. Returns:

• 200 OK if within rate limit
• 429 Too Many Requests if rate limit exceeded

Example response:

{
  "status": "ok",
  "key": "example-key",
  "rate_limit": "10/second",
  "timestamp": "2023-05-01T12:34:56.789012"
}

Configure Rate Limit

POST /configure

Configure a custom rate limit for a key.

Request body:

{
  "key": "example-key",
  "limit": "100/minute"
}

Response:

{
  "status": "ok",
  "key": "example-key",
  "rate_limit": "100/minute"
}

Get Statistics

GET /stats/{key}

Get hourly and daily usage statistics for a key.

Example response:

{
  "key": "example-key",
  "hourly_calls": {
    "2023-05-01-12": 45,
    "2023-05-01-11": 23
  },
  "daily_calls": {
    "2023-05-01": 152,
    "2023-04-30": 89
  },
  "current_rate_limit": "100/minute"
}

Health Check

GET /health

Check the health of the service.

Example response:

{
  "status": "ok",
  "timestamp": "2023-05-01T12:34:56.789012"
}

Environment Variables

• DEFAULT_RATE_LIMIT: Default rate limit (e.g., "10/second")
• PORT: Port to run the service on (default: 8000)

Running the Service

Using Docker

docker build -t rate-limiter .
docker run -p 8000:8000 rate-limiter

Using Python

pip install -r requirements.txt
python app.py

Rate Limit Format

Rate limits should be specified in the format {number}/{unit} where:

• number: A positive integer
• unit: One of "second", "minute", "hour", "day"

Examples:

• "10/second"
• "100/minute"
• "1000/hour"
• "5000/day"

Important Notes

This service uses in-memory storage, which means all rate limits and statistics will be lost when the service restarts. This is suitable for development and testing environments, but for production use cases requiring persistence, consider implementing a database backend.