Stream

About

The Stream is a high-performance HTTP-based transcoding service. Stream supports the TUS protocol, enabling reliable, resumable transcoding. Designed for seamless and consistent media processing, it supports advanced transcoding features with robust integration options.

---

Features

TUS Protocol Support

• Resumable transcoding: Leveraging the TUS protocol, Stream ensures reliable and efficient transcoding bucket processing.

Input Support

• Supported Input Formats:
  • mp4
  • webm

Output Options

• Supported Output Formats:
  • hls

Upload options

• s3 Upload: Direct upload to S3 storage.
• datalake Upload: Upload to datalake storage.

Key Functionalities

• Live transcoding with minimal upload time: Transcoding results are available after stream completion.
• Transcoding Cancellation: Cancel or pause ongoing transcoding in real-time.
• Transcoding Resumption: Resume incomplete transcoding tasks efficiently.

Transcoding scheduling

Installation

Prerequisites

• Go (v1.23+ recommended)
• ffmpeg (ensure it’s installed and available in your system's PATH)

Steps

1. Install dependencies:

go mod tidy

2. Build the service:

docker build . -t hcengineering/stream:latest

---

Configuration

App Env Configuration

The following environment variables can be used:

KEY                                  TYPE             DEFAULT                            DESCRIPTION
STREAM_LOG_LEVEL                     String           debug                              sets log level for the application
STREAM_SERVER_SECRET                 String                                              server secret required to generate and verify tokens
STREAM_PPROF_ENABLED                 True or False    true                               starts profile server on localhost:6060 if true
STREAM_INSECURE                      True or False    false                              ignores authorization check if true
STREAM_SERVE_URL                     String           0.0.0.0:1080                       listens on URL
STREAM_ENDPOINT_URL                  URL              s3://127.0.0.1:9000                S3 or Datalake endpoint, example: s3://my-ip-address, datalake://my-ip-address
STREAM_MAX_PARALLEL_SCALING_COUNT    Integer          2                                  how much parallel scaling can be processed
STREAM_MAX_THREAD_COUNT              Integer          4                                  max number of threads for transcoder
STREAM_OUTPUT_DIR                    String           /tmp/transcoding/                  path to the directory where transcoded files are stored
STREAM_SENTRY_DSN                    String           ""                                 sentry dsn for error tracking

Metadata

resolution: if passed, set the resolution for the output, for example, 'resolution: 1920:1080'.

token: must be provided to be authorized in the Huly's datalake service.

workspace: required for uploading content to the datalake storage.

S3 Env configuration

if you're working with S3 storage type, these envs must be provided: AWS_ACCESS_KEY_ID AWS_SECRET_ACCESS_KEY

Usage

The service exposes an HTTP API.

Below are some examples of how to interact with it.

Transcode via TUS

curl -X POST http://localhost:1080/recording \
     -H "Tus-Resumable: 1.0.0" \
     -H "Upload-Length: <file-size>" \
     --data-binary @path/to/your/file.mp4

Note: A TUS client is required to play with the service locally. You can use the tus-js-client example with video.

Schedule a transcoding

curl -X POST http://localhost:1080/transcoding \
  -H "Content-Type: application/json" \
  -H  "Authorization: Bearer <token>" \
  -d '{
      "source": "<input file name>",
      "format": "hls",
      "workspace": "test"
  }'

Contributing

We welcome contributions! To get started:

1. Fork the repository.
2. Create a new branch for your feature or bug fix.
3. Submit a pull request describing your changes.

---

License

This project is licensed under the MIT License.

---

Enjoy seamless transcoding with Stream! 🚀