Study of Real-Time Data Communication using Pulsar and RabbitMQ (Case Study of Stock Price in Lao Securities Exchange)

Abstract

This study aims to analyze the performance speed of data transmission between Pulsar and RabbitMQ, compare their efficiency in handling different message sizes ranging from 1 KB to 1,000 KB, and evaluate the suitability of each system for real-time data transmission in the Lao Securities Market. The testing was conducted in a Docker environment on MacOS Air M2 using NodeJS as the development tool. The experiment simulated real-time stock price data transmission in the Lao Securities Market, testing with progressively increasing message sizes.

The results showed that RabbitMQ performed faster with small-sized messages (1 - 10 KB), taking only 3 - 4.44 milliseconds compared to Pulsar's 16 - 17.7 milliseconds. However, when message sizes increased to 500 - 1,000 KB, Pulsar demonstrated superior performance maintenance, requiring only 19 - 26 milliseconds, while RabbitMQ's processing time significantly increased to 54 - 95 milliseconds.

These findings indicate that Pulsar is more suitable and efficient than RabbitMQ for securities trading systems that require handling larger message sizes and maintaining stable performance across varying data volumes. Meanwhile, RabbitMQ might be more appropriate for the high-speed transmission of smaller messages. The study's results provide valuable insights for selecting the appropriate messaging system based on specific use cases in the Lao Securities Market.