basic

WebRTC Video Call Example

This example demonstrates how to build a complete WebRTC video calling application using the Forge WebRTC extension.

Features

• Mesh and SFU Topologies: Support for peer-to-peer mesh calling and Selective Forwarding Unit (SFU) routing
• Quality Monitoring: Real-time connection quality tracking and adaptive bitrate
• Call Recording: Built-in recording capabilities with multiple formats
• Authentication: JWT-based authentication system
• Real-time Streaming: Integration with Forge streaming for presence and messaging
• Health Monitoring: Built-in health checks and metrics

Architecture

The example uses:

• WebRTC Extension: Core WebRTC functionality with mesh/SFU topologies
• Streaming Extension: Real-time presence, rooms, and messaging
• Auth Extension: JWT authentication and authorization
• HTTP API: REST endpoints for room management and call control

API Endpoints

Room Management

• POST /call/create - Create a new call room
• GET /call/{roomID} - Get room information and participants
• GET /calls - List all active call rooms

Call Operations

• POST /call/{roomID}/join - Join a call room
• POST /call/{roomID}/leave - Leave a call room
• GET /call/{roomID}/quality - Get call quality metrics

Recording

• POST /call/{roomID}/record/start - Start recording
• POST /call/{roomID}/record/stop - Stop recording

System

• GET /health - Health check endpoint
• WebSocket /webrtc/signal/{roomID} - WebRTC signaling endpoint

Configuration

The example is configured with:

WebRTC Settings

• Topology: SFU (Selective Forwarding Unit)
• STUN Servers: Google STUN servers for NAT traversal
• TURN Servers: Configurable relay servers
• Quality Monitoring: Enabled with 5-second intervals
• Recording: Enabled with WebM format

Streaming Settings

• Backend: Local (in-memory)
• Features: Rooms, channels, presence, typing indicators, message history
• Connection Limits: 5 connections per user
• Message Limits: 64KB max size, 100 messages per second

Usage

1. Build the example:

   cd examples/basic
   go build .
   

2. Run the server:

   ./basic
   

3. Create a call room:

   curl -X POST http://localhost:8080/call/create \
     -d "room_id=test-room" \
     -d "room_name=Test Room" \
     -d "max_members=10"
   

4. Join the call (with authentication):

   curl -X POST http://localhost:8080/call/test-room/join \
     -H "Authorization: Bearer YOUR_JWT_TOKEN" \
     -d "display_name=John Doe"
   

5. Check call quality:

   curl http://localhost:8080/call/test-room/quality
   

WebRTC Integration

The example includes a WebRTC signaling WebSocket endpoint at /webrtc/signal/{roomID}. Frontend clients should:

1. Connect to the signaling WebSocket
2. Exchange SDP offers/answers through the WebSocket
3. Use the configured STUN/TURN servers for ICE candidates
4. Implement video/audio capture and playback

Security Considerations

• Authentication is required for all call operations
• Room access is controlled through the streaming extension
• Configurable rate limiting prevents abuse
• Input validation on all endpoints

Development

The example demonstrates best practices for:

• Extension registration and dependency management
• HTTP route handling with proper error responses
• WebRTC configuration for production use
• Integration with other Forge extensions
• Health monitoring and observability

Scaling

For production deployment:

• Switch from local streaming backend to Redis/NATS
• Configure multiple TURN servers for global reach
• Implement load balancing for the SFU topology
• Add monitoring and alerting for call quality metrics