This project is a scalable video transcoding system built on AWS services and Node.js, aimed at automatically generating multiple resolutions of uploaded videos. Internally uses HLS (HTTP Live Streaming) and FFmpeg to generate multiple resolutions of the video.
Here is the backend demo of the video transcoding service:
- Automatic Video Transcoding: Automatically transcode the uploaded video into multiple resolutions.
- Scalable: The system is built on AWS services, making it scalable and reliable.
- Adaptive Bitrate Streaming: Uses HLS to generate multiple resolutions of the video for adaptive bitrate streaming.
- Dockerized: The system is dockerized for easy deployment and scaling.
The system is built on AWS services, including S3, Lambda, API Gateway, ECR and ECS. The architecture is as follows:
- User uploads a video to the S3 bucket.
- S3 triggers a Lambda function to transcode the video into multiple resolutions.
- The Lambda function sends the transcoded videos to another S3 bucket.
- Node.js
- AWS (S3, Lambda, API Gateway, ECR, ECS)
- Docker
- Unable to access the transcoded videos: The transcoded videos were not accessible due to incorrect permissions. This was resolved by updating the bucket policy to allow public access.
- Error while downloading video on docker: The video was not downloading on the docker container due to incorrect path. This was resolved by updating the path to the video in the code.
- Video transcoding taking too long: The transcoding process was taking too long due to the large size of the video. Use worker threads to improve the performance of the transcoding process.
- DRM (Digital Rights Management): Implement DRM to protect the content from unauthorized access.
- Live Streaming: Implement live streaming of videos using HLS.
- Automatically captions generation: Automatically generate captions for the videos using machine learning.
Feel free to contact me at email for any queries or suggestions.