Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/seatedro/drl
Distributed Rate Limiter
https://github.com/seatedro/drl
Last synced: about 2 months ago
JSON representation
Distributed Rate Limiter
- Host: GitHub
- URL: https://github.com/seatedro/drl
- Owner: seatedro
- License: mit
- Created: 2024-11-25T16:52:17.000Z (about 2 months ago)
- Default Branch: master
- Last Pushed: 2024-11-26T02:07:31.000Z (about 2 months ago)
- Last Synced: 2024-11-26T02:33:32.447Z (about 2 months ago)
- Language: Go
- Size: 1000 Bytes
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Distributed Rate Limiter (DRL)
A high-performance, distributed rate limiter built in Go, supporting both HTTP and gRPC protocols. Uses Redis for distributed state management and implements the Token Bucket algorithm for precise rate limiting.
## Features
- **Distributed Rate Limiting**: Consistent rate limiting across multiple service instances
- **Dual Protocol Support**: Both HTTP and gRPC interfaces
- **Token Bucket Algorithm**: Efficient and flexible rate limiting with burst support
- **Redis Backend**: Scalable and reliable state management
- **Configurable Limits**: Adjust rates, windows, and burst sizes via environment variables
- **Docker Ready**: Easy deployment with Docker and Docker Compose
## Architecture
```
[Client Requests]
↓
[Load Balancer]
↓
[Rate Limiter Service Instances] ←→ [Redis Cluster]
```
## Getting Started
### Prerequisites
- Go 1.22 or later
- Docker and Docker Compose
- Redis (for development)
### Installation
1. Clone the repository:
```bash
git clone https://github.com/seatedro/drl.git
cd drl
```
2. Install dependencies:
```bash
go mod download
```
3. Generate protocol buffers:
```bash
make generate
```
### Running with Docker Compose
```bash
docker-compose up --build
```
This will start:
- Rate limiter service (HTTP :8080, gRPC :9090)
- Redis instance
- Load testing service
## Usage
### HTTP API
```bash
# Check if request is allowed
curl -X POST "http://localhost:8080/v1/allow/my-key?namespace=myapp"
# Reset rate limit for a key
curl -X POST "http://localhost:8080/v1/reset/my-key?namespace=myapp"
```
### gRPC API
```bash
# Using grpcurl
grpcurl -d '{"key": "my-key", "namespace": "myapp"}' \
-plaintext localhost:9090 drl.v1.RateLimiter/Allow
```
### Response Headers (HTTP)
```
X-RateLimit-Remaining:
X-RateLimit-Reset:
Retry-After:
```
## Configuration
Environment variables:
| Variable | Description | Default |
|----------|-------------|---------|
| RATE_LIMIT | Requests allowed per window | 100 |
| RATE_WINDOW | Time window in seconds | 60 |
| BURST_SIZE | Maximum burst size | 150 |
| DEBUG | Enable debug logging | false |
## Load Testing
The project includes comprehensive load testing for both HTTP and gRPC endpoints:
```bash
# Run load tests
docker-compose up loadtest
```
Test scenarios include:
- Burst capacity testing
- Sustained load testing
- High concurrency testing
- Rate limit recovery testing