Skip to main content

System Architecture Overview

Envoy AI Gateway follows a modern cloud-native architecture pattern with distinct control and data planes. Let's explore how these components work together to manage AI traffic effectively.

High-Level Architecture​

The architecture is divided into two main planes:

  1. Control Plane: Responsible for configuring and managing the system

    • Kubernetes API Server serves as the configuration interface
    • AI Gateway Controller manages AI-specific resources and configurations
    • Envoy Gateway Controller handles core proxy configuration and xDS

  2. Data Plane: Handles the actual request traffic

    • Envoy Proxy and External Processor work together to handle AI traffic
    • Envoy Proxy routes and manages all incoming requests
    • External Processor handles AI-specific transformations and validations
    • Rate Limit Service manages token-based rate limiting

The Control Plane configures the Data Plane through several paths:

  • AI Gateway Controller configures the External Processor and Rate Limit Service
  • Envoy Gateway Controller configures Envoy Proxy through xDS

Component Interactions​

The system operates through a series of interactions between components:

  1. Configuration Flow

    • Kubernetes API Server receives configuration changes
    • AI Gateway Controller processes these changes and:
      • Creates resources for Envoy Gateway to consume
      • Configures the External Processor for AI request handling
      • Configures the Rate Limit Service for token-based limiting
    • Envoy Gateway Controller watches gateway resources and:
      • Translates them into Envoy configuration
      • Pushes updates to Envoy Proxy via xDS protocol

  2. Request Flow

    • Clients send requests to Envoy Proxy
    • Envoy Proxy routes requests through the External Processor for:
      • Request transformation and validation
      • Provider authentication
      • Token usage tracking
    • Processed requests are sent to AI providers
    • Responses flow back through the same path
    • Rate Limit Service enforces token-based limits

  3. Rate Limiting

    • External Processor extracts token usage from requests/responses
    • Rate Limit Service receives token usage data
    • Rate limits are enforced based on configured policies
    • Helps manage resource utilization and costs

Summary​

The Envoy AI Gateway architecture provides a robust system for managing AI traffic through:

  • Clear separation of control and data plane responsibilities
  • Flexible configuration management
  • Efficient request processing and transformation
  • Token-based rate limiting capabilities

For detailed information about specific components: