[mirotalksfu] - update config.js.template
هذا الالتزام موجود في:
Miroslav Pejic
@@ -1080,91 +1080,141 @@ module.exports = {
|
||||
* 🌐 Website : https://mediasoup.org
|
||||
*
|
||||
* 📚 Documentation:
|
||||
* ----------------
|
||||
* - Client API : https://mediasoup.org/documentation/v3/mediasoup-client/api/
|
||||
* - Server API : https://mediasoup.org/documentation/v3/mediasoup/api/
|
||||
* - Protocols : https://mediasoup.org/documentation/v3/mediasoup/rtp-parameters-and-capabilities/
|
||||
*
|
||||
* 🔧 Key Components:
|
||||
* -----------------
|
||||
* - Router : Manages RTP streams
|
||||
* - Transport : Network connection handler
|
||||
* - Producer : Media sender
|
||||
* - Consumer : Media receiver
|
||||
*
|
||||
* Mediasoup Configuration
|
||||
* -----------------------
|
||||
* This configuration defines settings for mediasoup workers, routers,
|
||||
* WebRTC servers, and transports. These settings control how the SFU
|
||||
* (Selective Forwarding Unit) handles media processing and networking.
|
||||
*/
|
||||
mediasoup: {
|
||||
/**
|
||||
* Worker Configuration
|
||||
* --------------------
|
||||
* - numWorkers: Number of mediasoup workers
|
||||
* - worker: Worker-specific settings
|
||||
* Workers are separate processes that handle media processing.
|
||||
* Multiple workers can run in parallel for load balancing.
|
||||
*/
|
||||
numWorkers: NUM_WORKERS,
|
||||
worker: {
|
||||
rtcMinPort: RTC_MIN_PORT,
|
||||
rtcMaxPort: RTC_MAX_PORT,
|
||||
rtcMinPort: RTC_MIN_PORT, // Minimum UDP/TCP port for ICE, DTLS, RTP
|
||||
rtcMaxPort: RTC_MAX_PORT, // Maximum UDP/TCP port for ICE, DTLS, RTP
|
||||
|
||||
// Disable Linux io_uring for certain operations (false = use if available)
|
||||
disableLiburing: false,
|
||||
|
||||
// Logging level (error, warn, debug, etc.)
|
||||
logLevel: process.env.MEDIASOUP_LOG_LEVEL || 'error',
|
||||
logTags: ['info', 'ice', 'dtls', 'rtp', 'srtp', 'rtcp', 'rtx', 'bwe', 'score', 'simulcast', 'svc', 'sctp'],
|
||||
|
||||
// Detailed logging for specific components:
|
||||
logTags: [
|
||||
'info', // General information
|
||||
'ice', // ICE (Interactive Connectivity Establishment) events
|
||||
'dtls', // DTLS handshake and encryption
|
||||
'rtp', // RTP packet flow
|
||||
'srtp', // Secure RTP encryption
|
||||
'rtcp', // RTCP control protocol
|
||||
'rtx', // Retransmissions
|
||||
'bwe', // Bandwidth estimation
|
||||
'score', // Network score calculations
|
||||
'simulcast', // Simulcast layers
|
||||
'svc', // Scalable Video Coding
|
||||
'sctp', // SCTP data channels
|
||||
],
|
||||
},
|
||||
numWorkers: NUM_WORKERS, // Number of mediasoup worker processes to create
|
||||
|
||||
/**
|
||||
* Router Configuration
|
||||
* --------------------
|
||||
* - Media codecs and capabilities
|
||||
* Routers manage media streams and define what codecs are supported.
|
||||
* Each mediasoup worker can host multiple routers.
|
||||
*/
|
||||
router: {
|
||||
// Enable audio level monitoring (for detecting who is speaking)
|
||||
audioLevelObserverEnabled: true,
|
||||
|
||||
// Disable active speaker detection (uses more CPU)
|
||||
activeSpeakerObserverEnabled: false,
|
||||
|
||||
/**
|
||||
* Supported Media Codecs
|
||||
* ----------------------
|
||||
* Defines what codecs the SFU can receive and forward.
|
||||
* Order matters - first is preferred during negotiation.
|
||||
*/
|
||||
mediaCodecs: [
|
||||
// Opus audio codec (standard for WebRTC)
|
||||
{
|
||||
kind: 'audio',
|
||||
mimeType: 'audio/opus',
|
||||
clockRate: 48000,
|
||||
channels: 2,
|
||||
clockRate: 48000, // Standard sample rate for WebRTC
|
||||
channels: 2, // Stereo audio
|
||||
},
|
||||
|
||||
// VP8 video codec (widely supported, good for compatibility)
|
||||
{
|
||||
kind: 'video',
|
||||
mimeType: 'video/VP8',
|
||||
clockRate: 90000,
|
||||
clockRate: 90000, // Standard video clock rate
|
||||
parameters: {
|
||||
'x-google-start-bitrate': 1000,
|
||||
'x-google-start-bitrate': 1000, // Initial bitrate (kbps)
|
||||
},
|
||||
},
|
||||
|
||||
// VP9 video codec (better compression than VP8)
|
||||
// Profile 0: Most widely supported VP9 profile
|
||||
{
|
||||
kind: 'video',
|
||||
mimeType: 'video/VP9',
|
||||
clockRate: 90000,
|
||||
parameters: {
|
||||
'profile-id': 0,
|
||||
'profile-id': 0, // Baseline profile
|
||||
'x-google-start-bitrate': 1000,
|
||||
},
|
||||
},
|
||||
|
||||
// VP9 Profile 2: Supports HDR and 10/12-bit color
|
||||
{
|
||||
kind: 'video',
|
||||
mimeType: 'video/VP9',
|
||||
clockRate: 90000,
|
||||
parameters: {
|
||||
'profile-id': 2,
|
||||
'profile-id': 2, // Advanced profile
|
||||
'x-google-start-bitrate': 1000,
|
||||
},
|
||||
},
|
||||
|
||||
// H.264 Baseline profile (widest hardware support)
|
||||
{
|
||||
kind: 'video',
|
||||
mimeType: 'video/h264',
|
||||
clockRate: 90000,
|
||||
parameters: {
|
||||
'packetization-mode': 1, // Required for WebRTC
|
||||
'profile-level-id': '42e01f', // Baseline 3.1
|
||||
'level-asymmetry-allowed': 1, // Allows different levels
|
||||
'x-google-start-bitrate': 1000,
|
||||
},
|
||||
},
|
||||
|
||||
// H.264 Main profile (better compression than Baseline)
|
||||
{
|
||||
kind: 'video',
|
||||
mimeType: 'video/h264',
|
||||
clockRate: 90000,
|
||||
parameters: {
|
||||
'packetization-mode': 1,
|
||||
'profile-level-id': '42e01f',
|
||||
'level-asymmetry-allowed': 1,
|
||||
'x-google-start-bitrate': 1000,
|
||||
},
|
||||
},
|
||||
{
|
||||
kind: 'video',
|
||||
mimeType: 'video/h264',
|
||||
clockRate: 90000,
|
||||
parameters: {
|
||||
'packetization-mode': 1,
|
||||
'profile-level-id': '4d0032',
|
||||
'profile-level-id': '4d0032', // Main 4.0
|
||||
'level-asymmetry-allowed': 1,
|
||||
'x-google-start-bitrate': 1000,
|
||||
},
|
||||
@@ -1175,23 +1225,54 @@ module.exports = {
|
||||
/**
|
||||
* WebRTC Server Configuration
|
||||
* ---------------------------
|
||||
* - Active: Enable/disable WebRTC server
|
||||
* - Options: Listening interfaces and ports
|
||||
* WebRTC servers handle ICE (connection establishment) and DTLS (encryption).
|
||||
* Can be disabled if using plain WebRtcTransport instead.
|
||||
*
|
||||
* Best used when:
|
||||
* - Running in controlled environments with fixed IPs
|
||||
* - Need to minimize port usage across workers
|
||||
* - Using StatefulSets/DaemonSets in Kubernetes
|
||||
*
|
||||
* Kubernetes considerations:
|
||||
* - Requires stable network identity (use StatefulSet)
|
||||
* - Needs NodePort/LoadBalancer with externalTrafficPolicy: Local
|
||||
* - Port ranges must be carefully allocated to avoid conflicts
|
||||
*/
|
||||
webRtcServerActive: process.env.SFU_SERVER === 'true',
|
||||
webRtcServerActive: process.env.SFU_SERVER === 'true', // Enable if SFU_SERVER=true
|
||||
webRtcServerOptions: {
|
||||
// Network interfaces and ports for ICE candidates
|
||||
listenInfos: [
|
||||
/**
|
||||
* UDP Configuration
|
||||
* Preferred for media transport (lower latency)
|
||||
* Kubernetes implications:
|
||||
* - Each Pod needs unique ports if sharing host network
|
||||
* - Consider using hostPort when not using LoadBalancer
|
||||
*/
|
||||
{
|
||||
protocol: 'udp',
|
||||
ip: LISTEN_IP,
|
||||
announcedAddress: IPv4,
|
||||
portRange: { min: RTC_MIN_PORT, max: RTC_MIN_PORT + NUM_WORKERS },
|
||||
ip: LISTEN_IP, // Local IP to bind to
|
||||
announcedAddress: IPv4, // Public IP sent to clients
|
||||
portRange: {
|
||||
min: RTC_MIN_PORT,
|
||||
max: RTC_MIN_PORT + NUM_WORKERS, // Port range per worker
|
||||
},
|
||||
},
|
||||
/**
|
||||
* TCP Configuration
|
||||
* Fallback for restrictive networks (higher latency)
|
||||
* Kubernetes implications:
|
||||
* - Helps with networks blocking UDP
|
||||
* - May require separate Service definition in k8s
|
||||
*/
|
||||
{
|
||||
protocol: 'tcp',
|
||||
ip: LISTEN_IP,
|
||||
announcedAddress: IPv4,
|
||||
portRange: { min: RTC_MIN_PORT, max: RTC_MIN_PORT + NUM_WORKERS },
|
||||
portRange: {
|
||||
min: RTC_MIN_PORT,
|
||||
max: RTC_MIN_PORT + NUM_WORKERS,
|
||||
},
|
||||
},
|
||||
],
|
||||
},
|
||||
@@ -1199,27 +1280,71 @@ module.exports = {
|
||||
/**
|
||||
* WebRTC Transport Configuration
|
||||
* ------------------------------
|
||||
* - Network and bandwidth settings
|
||||
* Transports handle the actual media flow between clients and the SFU.
|
||||
* These settings affect bandwidth management and network behavior.
|
||||
*
|
||||
* Preferred when:
|
||||
* - Running in cloud environments with auto-scaling
|
||||
* - Need dynamic port allocation
|
||||
* - Kubernetes Pods are ephemeral
|
||||
*
|
||||
* Kubernetes considerations:
|
||||
* - Requires wide port range exposure (50000-60000 typical)
|
||||
* - Works better with ClusterIP Services
|
||||
* - More resilient to Pod restarts
|
||||
*/
|
||||
webRtcTransport: {
|
||||
// Network interfaces for media transmission
|
||||
listenInfos: [
|
||||
/**
|
||||
* UDP Transport Settings
|
||||
* Kubernetes implications:
|
||||
* - Needs hostNetwork or privileged Pod for port access
|
||||
* - Consider port range size based on expected scale
|
||||
*/
|
||||
{
|
||||
protocol: 'udp',
|
||||
ip: LISTEN_IP,
|
||||
announcedAddress: IPv4,
|
||||
portRange: { min: RTC_MIN_PORT, max: RTC_MAX_PORT },
|
||||
portRange: {
|
||||
min: RTC_MIN_PORT,
|
||||
max: RTC_MAX_PORT, // Wider range than WebRtcServer
|
||||
},
|
||||
},
|
||||
/**
|
||||
* TCP Transport Settings
|
||||
* Kubernetes implications:
|
||||
* - Less efficient but more compatible
|
||||
* - May require different Service configuration
|
||||
*/
|
||||
{
|
||||
protocol: 'tcp',
|
||||
ip: LISTEN_IP,
|
||||
announcedAddress: IPv4,
|
||||
portRange: { min: RTC_MIN_PORT, max: RTC_MAX_PORT },
|
||||
portRange: {
|
||||
min: RTC_MIN_PORT,
|
||||
max: RTC_MAX_PORT,
|
||||
},
|
||||
},
|
||||
],
|
||||
initialAvailableOutgoingBitrate: 1000000,
|
||||
minimumAvailableOutgoingBitrate: 600000,
|
||||
maxSctpMessageSize: 262144,
|
||||
maxIncomingBitrate: 1500000,
|
||||
|
||||
/**
|
||||
* Bandwidth Control Settings
|
||||
* Kubernetes implications:
|
||||
* - These values should be tuned based on Node resources
|
||||
* - Consider network plugin overhead (Calico, Cilium etc.)
|
||||
*/
|
||||
initialAvailableOutgoingBitrate: 1000000, // 1 Mbps initial bitrate
|
||||
minimumAvailableOutgoingBitrate: 600000, // 600 Kbps minimum guaranteed
|
||||
maxIncomingBitrate: 1500000, // 1.5 Mbps max per producer
|
||||
|
||||
/**
|
||||
* Data Channel Settings
|
||||
* Kubernetes implications:
|
||||
* - Affects memory allocation per transport
|
||||
* - Larger sizes may require Pod resource adjustments
|
||||
*/
|
||||
maxSctpMessageSize: 262144, // 256 KB max message size for data channels
|
||||
},
|
||||
},
|
||||
};
|
||||
|
||||
المرجع في مشكلة جديدة
حظر مستخدم