Skip to main content

Create Rules

This guide will walk you through creating rules in AI Gateway. Rules determine how incoming requests are matched and routed to your services.

Prerequisites​

Before you begin, make sure you have:

  • AI Gateway installed and running
  • A service configured (see Add a Service)
  • Access to the Admin API

Step 1: Plan Your Rules​

Decide on your rule configuration:

  1. Choose the paths to match
  2. Determine allowed HTTP methods
  3. Plan path handling (strip path, preserve host)
  4. Identify target service

Step 2: Create a Rule​

Create your first rule using the Admin API:

curl -X POST http://localhost:8080/api/v1/gateways/{gateway-id}/rules \
  -H "Content-Type: application/json" \
  -d '{
    "service_id": "{service-id}",
    "path": "/v1/chat/completions",
    "methods": ["POST"],
    "strip_path": false,
    "preserve_host": false
  }'

Step 3: Add Path Matching​

Create rules with different path matching patterns:

# Prefix matching
curl -X POST http://localhost:8080/api/v1/gateways/{gateway-id}/rules \
  -H "Content-Type: application/json" \
  -d '{
    "service_id": "{service-id}",
    "path": "/api/v1/*",
    "methods": ["GET", "POST"],
    "strip_path": true
  }'

# Exact matching
curl -X POST http://localhost:8080/api/v1/gateways/{gateway-id}/rules \
  -H "Content-Type: application/json" \
  -d '{
    "service_id": "{service-id}",
    "path": "=/v1/models",
    "methods": ["GET"]
  }'

Step 4: Configure Headers​

Add header matching if needed:

curl -X POST http://localhost:8080/api/v1/gateways/{gateway-id}/rules \
  -H "Content-Type: application/json" \
  -d '{
    "service_id": "{service-id}",
    "path": "/v1/embeddings",
    "methods": ["POST"],
    "headers": {
      "content-type": ["application/json"]
    }
  }'

Step 5: Verify Rules​

Check that your rules are properly configured:

curl http://localhost:8080/api/v1/gateways/{gateway-id}/rules

Step 6: Test the Rules​

Test your rule configurations:

# Test a route
curl -X POST http://your-gateway-domain/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "x-api-key: your-key" \
  -d '{
    "model": "gpt-3.5-turbo",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Next Steps​

Now that you have configured your rules, read more about:

Troubleshooting​

Common issues and solutions:

  1. Routing Issues

    • Verify path patterns
    • Check method restrictions
    • Review rule priorities

  2. Path Handling

    • Check strip_path setting
    • Verify preserve_host
    • Test path transformations

  3. Header Problems

    • Verify header requirements
    • Check header values
    • Test header matching