Large-scale Distributed Storage System for Streaming Video Smoothly
Currently, on streaming video platforms, each video has hundreds of thousands of concurrent viewers, so the centralized storage system is unsuitable because the cost of high bandwidth and throughput is prohibitive. Therefore some researches focus on distributed storage. It has steadily grown in popularity, performance, stability, and scalability. The paper proposes an end-to-end distributed storage system for the streaming video industry. Our contribution includes the following: First, we offer a solution that allows personal computers to participate in a distributed storage system (storage pool) and connect to storage controllers. Next, an AI-based Chainflix Storage Controller algorithm is proposed that helps scale bandwidth and throughput for streaming video smoothly. Finally, we provide fault-tolerant solutions for the system when a node in the storage pool loses connection or when a node shows signs of overloading. The system works well on a Blockchain-based P2P Streaming Platform named Chainflix.
More detail about paper at the link: https://drive.google.com/file/d/18enAahF3lvLxL4aK_sI8gM3_3PoUNREr/view?usp=sharing