Golem Transcoding is a React + Spring based webapp accepting video files as input and transcoding these files into different formats using Golem.
Every input file will be converted to .mkv, .mov, .avi and .mp4 output files.
Here's a YouTube video explaining how the whole system works.
- docker - Ubuntu 18.04 base image with ffmpeg installed on top of it. Contains scripts to build the image and upload it to Yagna repository.
- requestor-python - Golem requestor script for splitting the work into subtasks and sending them to remote nodes.
- requestor-python/core.py - basic data types used across the project.
- requestor-python/engine.py - integration with Yagna API.
- requestor-python/requestor.py - entry point of the whole program.
- requestor-python/task.py - generation of subtask details.
- requestor-python/utils.py - utils copied from yapapi examples.
- requestor-python/create-venv.sh - creates virtual environment and installs dependencies.
- requestor-python/run.sh - executes the requestor.py script in the context of activated virtual environment.
- requestor-python/output - transcoded video file storage.
- requestor-spring - backend server based on Spring Boot. Serves frontend assets and integrates with requestor script to send tasks to Golem. Responsible for keeping information about the transcoding job status.
- requestor-spring/gradle - Gradle 6.7.1 distribution.
- requestor-spring/src/main/java - all Java classes consituting the backend part.
- requestor-spring/src/main/js - frontend part of the application. Webpack + React.
- requestor-spring/src/main/resources/public - static files served by the server.
- requestor-spring/src/main/resources/application.yml - Spring app configuration.
Golem setup. If you encounter an issue then refer to https://handbook.golem.network/requestor-tutorials/flash-tutorial-of-requestor-development for full set of instructions.
curl -sSf https://join.golem.network/as-requestor | bash -
~/.local/bin/yagna service run
/screen -S yagna -L -d -m ~/.local/bin/yagna service run
~/.local/bin/yagna payment init -r
yagna app-key create requestor
- Store the key for later.
App prerequisite
git clone https://github.com/Edhendil/golem-transcoding.git
Transcoding requestor setup
- Install python 3:
apt install python3
- Install pip:
apt install python3-pip
apt install python3-venv
cd golem-transcoding/requestor-python
./create-venv.sh
Server setup
- Setup JDK 8 using AdoptJDK website or with
apt install openjdk-8-jdk
. apt install python
(Python 2 is required for compilation of fibers packageapt install nodejs
cd golem-transcoding/requestor-spring
./gradlew :bootJar
cp build/libs/requestor-spring-1.0.0.jar ~/
(or any other location)cd ~
- Create
application.yml
file right next to therequestor-spring-1.0.0.jar
file - Set golem.scriptLocation to the absolute path of the requestor-python folder in the cloned repository. Example:
"/home/your-user/golem-transcoding/requestor-python"
. - Set golem.yagnaKey to the requestor key generated through the
yagna app-key create requestor
command. Example:"a0f1a0b7af0a42bcab0361f39dfd42b4"
- Set golem.inputFileLocation to the absolute path of the folder where you want to store the input video files. Example:
"/home/your-user/transcoding-inputs"
. Be sure to create such folder first. - Set golem.dataRetention to how long data about the jobs and their outputs should be stored. Example:
"1h"
. - Set golem.maxUpload to the max size of an accepted input file. KB, MB, GB suffixes are supported. Example:
"100MB"
- Final file should look like this:
golem: scriptLocation: "<absolute path to requestor script folder>" yagnaKey: "<Yagna requestor app key>" inputFileLocation: "<absolute path to the folder where input files will be stored>" dataRetention: "<how long should information about a job be kept in storage>" maxUpload: "<maximal size of the input file>"
java -jar requestor-spring-1.0.0.jar
/screen -S transcoding -L -d -m java -jar requestor-spring-1.0.0.jar
Access the webapp
- Open browser and go to
localhost:8080
MIT