Skip to main content
Version: 2.1.0

SFU

Example Configuration​

[log]
level = "info"

[nats]
url = "nats://127.0.0.1:4222"

[sfu]
ballast = 0

[router]
maxbandwidth = 1500
maxpackettrack = 500
audiolevelthreshold = 40
audiolevelinterval=1000
audiolevelfilter = 20

[router.simulcast]
bestqualityfirst = true

[webrtc]
singleport = 5000
sdpsemantics = "unified-plan"
mdns = true

[[webrtc.iceserver]]
urls = ["stun:stun.stunprotocol.org:3478"]

[[webrtc.iceserver]]
urls = ["turn:turn.awsome.org:3478"]
username = "awsome"
credential = "awsome"

[webrtc.candidates]
nat1to1 = ["127.0.0.1"]
icelite = true

[webrtc.timeouts]
disconnected = 5
failed = 25
keepalive = 2

log.level​

  • Type: string​

  • Default: 'info'​

  • The logging level. info for production, debug for development.

nats.url​

  • Type: string​

  • The URL of the NATS server.

sfu.ballast​

  • Type: int​

  • The ballast size in MiB, will allocate memory to reduce the GC trigger upto 2x the size of ballast. Be aware that the ballast should be less than the half of memory available.

router.maxbandwidth​

  • Type: int​

  • The maximum bandwidth in kbps. Zero for unlimited bandwidth.

router.maxpackettrack​

  • Type: int​

  • The maximum number of video packet tracks the SFU will handle.

router.audiolevelthreshold​

  • Type: int​

  • Sets the audio level volume threshold. Values from [0-127] where 0 is the loudest.

router.audiolevelinterval​

  • Type: int​

  • Sets the interval in which the SFU will check the audio level in [ms]. If the active speaker has changed, the SFU will emit an event to clients.

router.audiolevelfilter​

  • Type: int​

  • Sets minimum percentage of events required to fire an audio level according to the expected events from the audiolevelinterval, calculated as audiolevelinterval/packetization time (20ms for 8kHz). Values from [0-100].

router.simulcast.bestqualityfirst​

  • Type: bool​

  • If set to true, the SFU will prefer to send the highest quality layer first.

webrtc.singleport​

  • Type: int​

  • The port on which the SFU will listen for WebRTC connections.

webrtc.sdpsemantics​

  • Type: string​

  • The SDP semantics. unified-plan, plan-b or unified-plan-with-fallback.

webrtc.mdns​

  • Type: bool​

  • If set to true, the SFU will advertise itself via mDNS.

webrtc.iceserver.urls​

  • Type: string​

  • The URL of the ICE server.

webrtc.iceserver.username​

  • Type: string​

  • The username of the ICE server.

webrtc.iceserver.credential​

  • Type: string​

  • The credential of the ICE server.

webrtc.candidates.nat1to1​

  • Type: []string​

  • In case you're deploying ion-sfu on a server which is configured with a 1:1 NAT (e.g., Amazon EC2), you might want to also specify the public address of the machine using the setting below. This will result in all host candidates (which normally have a private IP address) to be rewritten with the public address provided in the settings. As such, use the option with caution and only if you know what you're doing. Multiple public IP addresses can be specified as a comma separated list if the sfu is deployed in a DMZ between two 1-1 NAT for internal and external users.

webrtc.candidates.icelite​

  • Type: bool​

  • If set to true, the SFU will use ICE-lite.

webrtc.timeouts.disconnected​

  • Type: int​

  • The duration in [sec] without network activity before a ICE Agent is considered disconnected.

webrtc.timeouts.failed​

  • Type: int​

  • The duration in [sec] without network activity before a ICE Agent is considered failed after disconnected.

webrtc.timeouts.keepalive​

  • Type: int​

  • How often in [sec] the ICE Agent sends extra traffic if there is no activity, if media is flowing no traffic will be sent.