A self contained OBS -> FTL -> WebRTC live streaming server. Comprised of 3 parts once configured anyone can achieve sub-second OBS to the browser livestreaming
View Demo
·
Report a Bug
·
Request Features
Project Lightspeed is a fully self-contained live streaming server. With Lightspeed you will be able to deploy your own sub-second latency live streaming platform. The Lightspeed repository contains the instructions for installing and deploying the entire application. So far, Lightspeed includes an ingest service, broadcast service via webRTC and a web application for viewing. Lightspeed is however completely modular. What this means is that you can write your own web app, ingest server or broadcast server.
Lightspeed Ingest listens on port 8084 which is the port used by the FTL protocol. Upon receiving a connection it completes the FTL handshake and negotiates a port (this is currently bugged however and defaults to 65535). Once the negotiation is done Lightspeed WebRTC listens on the negotiated port (in the future Lightspeed WebRTC will listen on the loopback interface so the ingest has more control on what packets we accept) and relays the incoming RTP packets over WebRTC. Lightspeed React communicates via websocket with Lightspeed WebRTC to exchange ICE Candidates and once a connection is established the video can be viewed.
Here is a diagram that outlines the current implementation and the future implementation that I would like to achieve. The reason I want the packets relayed from Ingest to WebRTC on the loopback interface is so that we have more control over who can send packets. Meaning that when a DISCONNECT command is recieved we can terminate the UDP listener so that someone could not start sending packets that we do not want
- Rust
- Golang
- React
We now have a Discord server! This is a great way to stay up to date with the project and join in on the conversation! Come stop by!
In order to get a copy running you will need to install all 3 repositories. There are installation instructions in each repo however I will include them here for the sake of simplicity.
In order to run Lightspeed, Golang, Rust, and npm are required. Additionally the Rust repo requires a C compiler. If you get a linker cc not found error then you need to install a C compiler.
git clone https://github.com/GRVYDEV/Lightspeed-ingest.git
cd Lightspeed-ingest
cargo buildUsing go get
export GO111MODULE=on
go get github.com/GRVYDEV/lightspeed-webrtcUsing git
git clone https://github.com/GRVYDEV/Lightspeed-webrtc.git
cd Lightspeed-webrtc
export GO111MODULE=on
go buildgit clone https://github.com/GRVYDEV/Lightspeed-react.git
cd Lightspeed-react
npm installSome of our awesome community members have written their own installers for Lightspeed. Here are links to those!
Note: If you want to make a custom installer do so in the /contrib folder and submit a PR. Please make sure to include a README on how to use it!
cd Lightspeed-ingest
cargo run --releaseUsing go get
lightspeed-webrtc --addr=XXX.XXX.XXX.XXXUsing git
cd Lightspeed-webrtc
go build
./lightspeed-webrtc --addr=XXX.XXX.XXX.XXX| Argument | Supported Values | Defaults | Notes |
|---|---|---|---|
--addr |
A valid IP address | localhost |
This is the local Ip address of your machine. It defaults to localhost but should be set to your local IP. For example 10.17.0.5 This is where the server will listen for UDP packets and where it will host the websocket endpoint for SDP negotiation |
--ip |
A valid IP address | none |
Sets the public IP address for WebRTC to use. This is especially useful in the context of Docker |
--ports |
A valid UDP port range | 20000-20500 |
This sets the UDP ports that WebRTC will use to connect with the client |
--ws-port |
A valid port number | 8080 |
This is the port on which the websocket will be hosted. If you change this value make sure that is reflected in the URL used by the react client |
--rtp-port |
A valid port number | 65535 |
This is the port on which the WebRTC service will listen for RTP packets. Ensure this is the same port that Lightspeed Ingest is negotiating with the client |
You should then configure the websocket URL in config.json in the build directory. If you are using an IP then it will be the
public IP of your machine if you have DNS then it will be your hostname.
Note: The websocket port is hardcoded meaning that Lightspeed-webrtc will always serve it on port 8080 (this may change in the future)
so for the websocket config it needs to be ws://IP_or_Hostname:8080/websocket
You can host the static site locally using serve which can be found here
Note: your version of serve may require the -p flag instead of -l for the port
cd Lightspeed-react
npm run build
serve -s build -l 80The above will serve the build folder on port 80.
View Lightspeed in your web browser by visiting http://hostname or http://your.ip.address.here
Install Docker and docker-compose.
See the .env file to configure per your needs. At minimum, you need to set WEBSOCKET_HOST. The stream key will be
generated automatically on boot, and change each restart, unless you set a static one.
Use docker-compose up to start all containers at once and monitor the logs. When you are happy it is working you can
move to running detached.
Use docker-compose up -d to start all containers detached to have them run in the background.
Use docker ps to verify uptime, port forwarding, etc.
You can also use docker-compose logs -f to follow the logs of all the containers, and press CTRL + C to stop
following but leave the containers running.
For development purposes you can choose to build the containers locally instead of Docker Hub. Uncomment the build:
in the docker-compose.yaml. Configure the context: to be the path where you have cloned the respective respository
(React, WebRTC, or
Ingest). For example, create a base folder and clone each repostiory
there as such:
mkdir Lightspeed
git clone ...
git clone ...
git clone ...
---
./Lightspeed # base folder
Project-Lightspeed/ # you are here
Lightspeed-react/
Lightspeed-ingest/
Lightspeed-webrtc/
Run docker-compose build to build the local container images. If you change the source code you will need to run again.
You can run rebuild an individual container via docker-compose build lightspeed-react.
By default, since we are using the FTL protocol, you cannot just use a Custom server. You will need to edit
your services.json file. It can be found at:
- Windows:
%AppData%\obs-studio\plugin_config\rtmp-services\services.json - OSX:
/Users/YOURUSERNAME/Library/Application\ Support/obs-studio/plugin_config/rtmp-services/services.json
Note: Not all versions of Linux have access to OBS with the FTL SDK built in. If you are on Linux and you cannot stream to Lightspeed this may be the issue.
Paste the below into the services array and change the url to either the IP or the hostname of your Project Lightspeed server
Note: for the url it is not prefaced by anything. For example, given an IP of 10.0.0.2 you would put "url": "10.0.0.2" You do not need to indicate a port since the FTL protocol always uses 8084
{
"name": "Project Lightspeed",
"common": false,
"servers": [
{
"name": "SERVER TITLE HERE",
"url": "your.lightspeed.hostname"
}
],
"recommended": {
"keyint": 2,
"output": "ftl_output",
"max audio bitrate": 160,
"max video bitrate": 8000,
"profile": "main",
"bframes": 0
}
},NOTE: You do not need to specify a port.
After restarting OBS you should be able to see your service in the OBS settings Stream pane. (Special Thanks to Glimesh for these instructions)
We are no longer using a default streamkey! If you are still using one please pull from master on the Lightspeed-ingest
repository. Now, by default on first time startup a new streamkey will be generated and output to the terminal for you.
In order to regenerate this key simply delete the file it generates called hash. In a Docker context we will work to
make the key reset process as easy as possible. Simply copy the key output in the terminal to OBS and you are all set!
This key WILL NOT change unless the hash file is deleted.
This project is still very much a work in progress and a lot of improvements will be made to the deployment process. If something is unclear or you are stuck there are two main ways you can get help.
- Discord - this is the quickest and easiest way I will be able to help you through some deployment issues.
- Create an Issue - this is another way you can bring attention to something that you want fixed.
I will be fleshing out the roadmap in the coming days. As of right now I want to get this to a point where it is as close to other live streaming services as possible. If there are any features that you want to see then feel free to suggest them!
See the open issues for a list of proposed features (and known issues).
I am very from perfect and there are bound to be bugs and things I've overlooked in the installation process. Please, add issues and feel free to reach out if anything is unclear. Also, we have a Discord.
Contributions are what make the open source community such an amazing place to be learn, inspire, and create. Any contributions you make are greatly appreciated.
- Fork the Project
- Create your Feature Branch:
git checkout -b feature/AmazingFeature - Commit your Changes:
git commit -m 'Add some AmazingFeature' - Push to the Branch:
git push origin feature/AmazingFeature - Open a Pull Request
Distributed under the MIT License. See LICENSE for more information.
Garrett Graves - @grvydev
Project Link: https://github.com/GRVYDEV/Project-Lightspeed
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent