Real-Time Video Streaming on the Pub/Sub Architecture: Case of Apache Kafka

The Pub/Sub architecture, widely employed for event-driven applications like IoT and complex event processing, faces challenges in its raw application to video streaming. This study investigates the specific operational demands inherent in implementing real-time video streaming within the Pub/Sub architecture, focusing on the Apache Kafka framework. The analysis delves into crucial aspects such as the relationship between brokers and partitions, the influence of compression on throughput, and the incorporation of core implementation techniques like multithreading, all crucial tools for ensuring sustainable and reliable video streaming. Additionally, this research utilizes the SINETStream library, designed for Apache Kafka, the SINETStream is a software package that makes it easy for researchers to reliably collect and analyze data via wide area networks without loss, enhancing adaptability and refinement. Through these investigations, the study contributes nuanced insights into the considerations essential for optimizing video streaming within the Pub/Sub architecture, thereby advancing the understanding and practical implementation of such systems.