send me the egress.yaml configuration for self hosting

Thanks for reaching out, Youssef. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

S3 , plain Docker Compose,

Thanks — I can create the egress.yaml for S3 + plain Docker Compose. I just need a few small details so the config matches your environment. Could you confirm the following? • LiveKit version you’re running (or say “latest” if unsure). • Is the S3 provider AWS S3 or an S3-compatible service (e.g., MinIO)? If compatible, what's the endpoint URL? • S3 region and bucket name (you can redact actual names if you prefer). • Do you need path-style addressing (

forcePathStyle: true

) or virtual-hosted–style? • How will you provide credentials: environment variables (access key / secret) or an IAM role? (You can redact the secret values.) • Which output formats do you want saved to S3 (e.g.,

mp4

,

webm

,

hls

)? Any preferred filename or prefix pattern (for example:

recordings/{room}/{time}.mp4

)? • Should uploaded objects be public-read or private? • Where will this

egress.yaml

be mounted/loaded from in your Docker Compose setup? (path in the host/container) • Any TLS/custom CA required for the S3 endpoint? If you prefer, give me the common defaults and I’ll produce a complete example (Docker Compose +

egress.yaml

) using: • AWS S3 • region

us-east-1

• bucket

my-livekit-recordings

•

forcePathStyle: false

• credentials via environment variables • output

mp4

files saved under

recordings/{room}/{start_time}.mp4

• private ACL Tell me which of the defaults to use or send the values above, and I’ll generate the ready-to-use files. , with interactive elements

latest, AWS S3, .ogg, access key, secret, private,

Thanks — I’ve prepared a ready-to-use example for a self-hosted LiveKit egress configuration that writes .ogg recordings to AWS S3, plus a minimal Docker Compose snippet that mounts the config and supplies credentials via environment variables. Notes: • This uses LiveKit “latest” image per your request. • Files are uploaded with private ACL. • Credentials are passed via environment variables (redact real secrets in examples). • Adjust bucket/region/prefix values to match your environment. 1) egress.yaml

# egress.yaml - S3 output, .ogg recordings
egress:
  outputs:
    - name: s3-ogg
      type: s3
      s3:
        bucket: my-livekit-recordings       # replace with your bucket name
        region: us-east-1                   # replace if different
        prefix: recordings/{room}/{start_time}  # path/prefix pattern
        acl: private
        force_path_style: false             # set true for MinIO or path-style endpoints
        content_type: audio/ogg
        # Optional: set server_side_encryption: AES256
  recording:
    # default recording options
    audio:
      enabled: true
      format: ogg
    video:
      enabled: false
    # If you want combined audio+video later, enable video and change format accordingly
  filename:
    # fallback filename template for outputs that use it
    template: "{room}-{start_time}.ogg"

2) Docker Compose snippet

version: "3.8"
services:
  livekit:
    image: livekit/livekit-server:latest
    container_name: livekit
    restart: unless-stopped
    ports:
      - "7880:7880"   # API/HTTP if used
      - "7881:7881"   # optional admin / metrics (adjust as needed)
      - "7882:7882/udp" # (example) media ports - ensure your networking matches LiveKit requirements
    environment:
      - LIVEKIT_KEY=your_api_key_here
      - LIVEKIT_SECRET=your_api_secret_here
      - S3_ACCESS_KEY_ID=${S3_ACCESS_KEY_ID}     # set in host env or replace inline
      - S3_SECRET_ACCESS_KEY=${S3_SECRET_ACCESS_KEY}
      - AWS_REGION=us-east-1
      # any other LiveKit environment variables you need
    volumes:
      - ./egress.yaml:/etc/livekit/egress.yaml:ro
    command:
      - "--egress-config=/etc/livekit/egress.yaml"

3) Environment / secrets • Recommended: set environment variables on the host or in a secure secrets manager and reference them in Compose. Example .env:

S3_ACCESS_KEY_ID=AKIA...
S3_SECRET_ACCESS_KEY=abcd...

4) How this works / quick tips • The

egress.yaml

defines an S3 output named

s3-ogg

that saves audio as .ogg. Change

bucket

,

region

, or

prefix

to match your needs. • The

prefix

supports simple templates (I used

{room}

and

{start_time}

as examples). If you want a different pattern, update the

prefix

and

filename.template

. • If you use a custom S3 endpoint (MinIO or other), set

force_path_style: true

and provide an

endpoint

field under

s3:

and ensure TLS/CA settings if needed. • Keep your AWS keys private. Using IAM role or instance profile is more secure if you later move to EC2/ECS. If you want, I can: • Replace the placeholder values with your actual bucket name and region (you can redact secrets). • Add an example of how to trigger an egress (API snippet) to generate .ogg recordings. • Convert this to a more complete Docker Compose that includes recommended LiveKit ports and TURN configuration. Which of those would you like next? , with interactive elements