mboogert/jitsi-web repository overview

⁠Jitsi Meet on Docker

Jitsi⁠ is a set of Open Source projects that allows you to easily build and deploy secure videoconferencing solutions.

Jitsi Meet⁠ is a fully encrypted, 100% Open Source videoconferencing solution that you can use all day, every day, for free — with no account needed.

This repository contains the necessary tools to run a Jitsi Meet stack on Docker⁠ using Docker Compose⁠.

⁠Table of contents

• Quick start⁠
• Architecture⁠
  • Images⁠
  • Design considerations⁠
• Configurations⁠
  • Advanced configuration⁠
  • Running on a LAN environment⁠
• Limitations⁠

⁠Quick start

In order to quickly run Jitsi Meet on a machine running Docker and Docker Compose, follow these steps:

• Clone this repository to your own computer.
  • git clone https://github.com/jitsi/docker-jitsi-meet && cd docker-jitsi-meet
• Create a .env file by copying and adjusting env.example, and create required CONFIG directories
  • cp env.example .env
  • mkdir -p ~/.jitsi-meet-cfg/{web/letsencrypt,transcripts,prosody,jicofo,jvb}
• Run docker-compose up -d.
• Access the web UI at https://localhost:8443⁠ (or a different port, in case you edited the compose file).

Note that HTTP (not HTTPS) is also available (on port 8000, by default), but that's e.g. for a reverse proxy setup; direct access via HTTP instead HTTPS leads to WebRTC errors such as Failed to access your microphone/camera: Cannot use microphone/camera for an unknown reason. Cannot read property 'getUserMedia' of undefined or navigator.mediaDevices is undefined.

If you want to use jigasi too, first configure your env file with SIP credentials and then run Docker Compose as follows: docker-compose -f docker-compose.yml -f jigasi.yml up

If you want to enable document sharing via Etherpad⁠, configure it and run Docker Compose as follows: docker-compose -f docker-compose.yml -f etherpad.yml up

If you want to use jibri too, first configure host as described in JItsi BRoadcasting Infrastructure configuration section and then run Docker Compose as follows: docker-compose -f docker-compose.yml -f jibri.yml up -d or to use jigasi too: docker-compose -f docker-compose.yml -f jigasi.yml -f jibri.yml up -d

⁠Architecture

A Jitsi Meet installation can be broken down into the following components:

• A web interface
• An XMPP server
• A conference focus component
• A video router (could be more than one)
• A SIP gateway for audio calls
• A Broadcasting Infrastructure for recording or streaming a conference.

The diagram shows a typical deployment in a host running Docker. This project separates each of the components above into interlinked containers. To this end, several container images are provided.

⁠External Ports

The following external ports must be opened on a firweall:

• 80/tcp for Web UI HTTP (really just to redirect, after uncommenting ENABLE_HTTP_REDIRECT=1 in .env)
• 443/tcp for Web UI HTTPS
• 4443/tcp for RTP media over TCP
• 10000/udp for RTP media over UDP

Also 20000-20050/udp for jigasi, in case you choose to deploy that to facilitate SIP acces.

E.g. on a CentOS/Fedora server this would be done like this (without SIP access):

    $ sudo firewall-cmd --permanent --add-port=80/tcp
    $ sudo firewall-cmd --permanent --add-port=443/tcp
    $ sudo firewall-cmd --permanent --add-port=4443/tcp
    $ sudo firewall-cmd --permanent --add-port=10000/udp
    $ sudo firewall-cmd --reload

⁠Images

• base: Debian stable base image with the S6 Overlay⁠ for process control and the Jitsi repositories⁠ enabled. All other images are based off this one.
• base-java: Same as the above, plus Java (OpenJDK).
• web: Jitsi Meet web UI, served with nginx.
• prosody: Prosody⁠, the XMPP server.
• jicofo: Jicofo⁠, the XMPP focus component.
• jvb: Jitsi Videobridge⁠, the video router.
• jigasi: Jigasi⁠, the SIP (audio only) gateway.
• etherpad: Etherpad⁠, shared document editing addon.
• jibri: Jibri⁠, the brooadcasting infrastructure.

⁠Design considerations

Jitsi Meet uses XMPP for signalling, thus the need for the XMPP server. The setup provided by these containers does not expose the XMPP server to the outside world. Instead, it's kept completely sealed, and routing of XMPP traffic only happens on a user defined network.

The XMPP server can be exposed to the outside world, but that's out of the scope of this project.

⁠Configuration

The configuration is performed via environment variables contained in a .env file. You can copy the provided env.example file as a reference.

IMPORTANT: At the moment, configuration is not regenerated on every container boot, so if you make any changes to your .env file, make sure you remove the configuration directory before starting your containers again.

NOTE: The mobile apps won't work with self-signed certificates (the default) see below for instructions on how to obtain a proper certificate with Let's Encrypt.

⁠Let's Encrypt configuration

If you plan on exposing this container setup to the outside traffic directly and want a proper TLS certificate, you are in luck because Let's Encrypt support is built right in. Here are the required options:

In addition, you will need to set HTTP_PORT to 80 and HTTPS_PORT to 443.

⁠SIP gateway configuration

If you want to enable the SIP gateway, these options are required:

⁠JItsi BRoadcasting Infrastructure configuration

Before running Jibri, you need to setup an ALSA loopback device on the host. This will not work on a non-Linux host.

For CentOS 7, the module is already compiled with the kernel, so just run:

# configure 5 capture/playback interfaces
echo "options snd-aloop enable=1,1,1,1,1 index=0,1,2,3,4" > /etc/modprobe.d/alsa-loopback.conf
# setup autoload the module
echo "snd_aloop" > /etc/modules-load.d/snd_aloop.conf
# load the module
modprobe snd-aloop
# check that the module is loaded
lsmod | grep snd_aloop

For Ubuntu:

# install the module
apt update && apt install linux-image-extra-virtual
# configure 5 capture/playback interfaces
echo "options snd-aloop enable=1,1,1,1,1 index=0,1,2,3,4" > /etc/modprobe.d/alsa-loopback.conf
# setup autoload the module
echo "snd-aloop" >> /etc/modules
# check that the module is loaded
lsmod | grep snd_aloop

NOTE: if you are running on AWS you may need to reboot your machine to use the generic kernel instead of the "aws" kernel.

If you want to enable Jibri these options are required:

Extended Jibri configuration:

For using multiple Jibri instances, you have to select different loopback interfces for each instance manually.

...
slave.pcm "hw:Loopback,0,0"
...
slave.pcm "hw:Loopback,0,1"
...
slave.pcm "hw:Loopback,1,1"
...
slave.pcm "hw:Loopback,1,0"
...

...
slave.pcm "hw:Loopback_1,0,0"
...
slave.pcm "hw:Loopback_1,0,1"
...
slave.pcm "hw:Loopback_1,1,1"
...
slave.pcm "hw:Loopback_1,1,0"
...

...
slave.pcm "hw:2,0,0"
...
slave.pcm "hw:2,0,1"
...
slave.pcm "hw:2,1,1"
...
slave.pcm "hw:2,1,0"
...

⁠Authentication

Authentication can be controlled with the environment variables below. If guest access is enabled, unauthenticated users will need to wait until a user authenticates before they can join a room. If guest access is not enabled, every user will need to authenticate before they can join.

⁠Internal authentication

The default authentication mode (internal) uses XMPP credentials to authenticate users. To enable it you have to enable authentication with ENABLE_AUTH and set AUTH_TYPE to internal, then configure the settings you can see below.

Internal users must be created with the prosodyctl utility in the prosody container. In order to do that, first execute a shell in the corresponding container:

docker-compose exec prosody /bin/bash

Once in the container, run the following command to create a user:

prosodyctl --config /config/prosody.cfg.lua register username meet.jitsi passsword

The command then asks for a password interactively.

⁠Authentication using LDAP

You can use LDAP to authenticate users. To enable it you have to enable authentication with ENABLE_AUTH and set AUTH_TYPE to ldap, then configure the settings you can see below.

⁠Authentication using JWT tokens

You can use JWT tokens to authenticate users. To enable it you have to enable authentication with ENABLE_AUTH and set AUTH_TYPE to jwt, then configure the settings you can see below.

This can be tested using the jwt.io⁠ debugger. Use the following samople payload:

{
  "context": {
    "user": {
      "avatar": "https://robohash.org/john-doe",
      "name": "John Doe",
      "email": "[email protected]"
    }
  },
  "aud": "my_jitsi_app_id",
  "iss": "my_jitsi_app_id",
  "sub": "meet.jitsi",
  "room": "*"
}

⁠Shared document editing using Etherpad

You can collaboratively edit a document via Etherpad⁠. In order to enable it, set the config options bellow and run Docker Compose with the additional config file etherpad.yml.

Here are the required options:

⁠Transcription configuration

If you want to enable the Transcribing function, these options are required:

For setting GOOGLE_APPLICATION_CREDENTIALS please read https://cloud.google.com/text-to-speech/docs/quickstart-protocol⁠ section "Before you begin" from 1 to 5 paragraph.

⁠Advanced configuration

These configuration options are already set and generally don't need to be changed.

⁠Running behind NAT or on a LAN environment

If running in a LAN environment (as well as on the public Internet, via NAT) is a requirement, the DOCKER_HOST_ADDRESS should be set. This way, the Videobridge will advertise the IP address of the host running Docker instead of the internal IP address that Docker assigned it, thus making ICE⁠ succeed.

The public IP address is discovered via STUN⁠. STUN servers can be specified with the JVB_STUN_SERVERS option.

⁠TODO

• Support container replicas (where applicable).
• TURN server.