Skip to content

Latest commit

 

History

History
110 lines (79 loc) · 2.66 KB

readme.md

File metadata and controls

110 lines (79 loc) · 2.66 KB
Raw

Livestreaming / Recording using Xvfb, ffmpeg, and PulseAudio.

System works very simply, it uses Xvfb to create a virtual display, PulseAudio to create a virtual audio device, and ffmpeg to record the screen and audio. Setups chromedp for the browser automation to record any browser url.

Requirements

  • docker.

Usage

docker build -t recorder-dev .

Preffred to use the docker-compose to run the container.

docker-compose up

SSH into the container to run the server.

make run

NOTE: Check you are inside the app working directory.

Check the server is running by visiting the http://localhost:3000/.

Sometimes Pulse Audio just doesn't want to start, so to give it a little push,
After you SSH inside the container, run the following command.
`./pulseaudio.sh`

ENVIRONMENT VARIABLES

  • BUCKET_ENDPOINT - AWS S3 Bucket Endpoint.
  • BUCKET_NAME - AWS S3 Bucket Name.
  • BUCKET_KEY_ID - AWS S3 Bucket Key ID.
  • BUCKET_APP_KEY - AWS S3 Bucket Secret Key.
  • BUCKET_REGION - AWS S3 Bucket Region.

API ENDPOINTS

  • /ping - To check the server is running.
curl --location 'http://localhost:3000/ping'
  • /start-recording - To start the recording.
curl --location 'http://localhost:3000/start-recording' \
--header 'Content-Type: application/json' \
--data '{
    "record_url": "https://www.youtube.com/watch?v=cii6ruuycQA&ab_channel=OliviaRodrigoVEVO"
    "stream_url: "rtmp://a.rtmp.youtube.com/live2/<stream_key>"
}'
  • /stop-recording - To stop the recording. Use the id from the start-recording response.
curl --location --request PATCH 'http://localhost:3000/stop-recording' \
--header 'Content-Type: application/json' \
--data '{
    "id": "pipeline_1725213615468"
}'

TODO

  • Add API server Capabilties to make custom recording calls.
  • Ability to record multiple recordings at the same time. ( Took some time to figure out the pulse audio thing )
  • Do proper file system implementation to handle different recordings.
  • Add Capabilties to do livestreaming onto multiple rtmp, rtmps endpoints ( Hopefully ffmpeg will make it easy )
  • Maybe do the kubernetes thing for auto scaling.
  • Benchmarking, because I know sometimes recordings will fail to record propely, because its life.
  • Add more documentation.
  • Add more tests

Scripts

  • To get all the running pulse audio
pactl list short sources
  • To Close the pulse audio
pulseaudio -k
  • To Close all the Pulse Sink
pactl unload-module module-null-sink
  • To Read all the running Process
ps aux

very useful to check for memory leaks, and zombie processes.