Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.draskencloud.com/llms.txt

Use this file to discover all available pages before exploring further.

Rate limiting is configured via the rate_limit field in any policy. It applies after authentication and before proxying.

​
Basic Configuration

{
  "rate_limit": {
    "enabled": true,
    "requests": 1000,
    "window_seconds": 60,
    "algorithm": "fixed_window"
  }
}

​
Fields

FieldDefaultDescription
enabledfalseEnable rate limiting
requests—Maximum requests allowed per window
window_seconds—Window duration in seconds
algorithmfixed_windowAlgorithm to use (see below)

​
Algorithms

AlgorithmDescription
fixed_windowResets the counter at fixed intervals. Fast, simple, but allows bursts at window boundaries
sliding_windowSmooths out bursts by tracking a rolling time window
token_bucketAllows short bursts up to a capacity, then enforces a steady refill rate

​
Applying at Different Levels

​
App-level (applies to all requests)

curl -X POST http://localhost:9090/api/v1/apps \
  -d '{
    "name": "My API",
    "slug": "my-api",
    "gateway_path": "/my-api",
    "default_policy": {
      "rate_limit": {
        "enabled": true,
        "requests": 5000,
        "window_seconds": 60
      }
    }
  }'

​
Endpoint-level override (stricter on a specific route)

{
  "policy_override": {
    "rate_limit": {
      "enabled": true,
      "requests": 10,
      "window_seconds": 60,
      "algorithm": "sliding_window"
    }
  }
}

​
Global Rate Limit

A global default can be set via environment variables: 
GATEWAY_RATE_LIMIT_REQUESTS=1000
GATEWAY_RATE_LIMIT_WINDOW_SECS=60

​
Per-IP Rate Limiting

Enable per-IP rate limiting globally: 
GATEWAY_PER_IP_RATE_LIMIT_ENABLED=true
GATEWAY_PER_IP_RATE_LIMIT_REQUESTS=100
GATEWAY_PER_IP_RATE_LIMIT_WINDOW_SECS=60
GATEWAY_TRUSTED_PROXY_IPS=10.0.0.1,10.0.0.2
Set GATEWAY_TRUSTED_PROXY_IPS when behind a load balancer so the real client IP is extracted from X-Forwarded-For.

​
Rate Limit Response

When a client exceeds the limit, the gateway returns: 
HTTP/1.1 429 Too Many Requests
Retry-After: 60