Benchmarking Message Queues

Abstract

Message queues are a way for different software components or applications to communicate with each other asynchronously by passing messages through a shared buffer. This allows a sender to send a message without needing to wait for an immediate response from the receiver, which can help to improve the system’s performance, reduce latency, and allow components to operate independently. In this paper, we compared and evaluated the performance of four popular message queues: Redis, ActiveMQ Artemis, RabbitMQ, and Apache Kafka. The aim of this study was to provide insights into the strengths and weaknesses of each technology and to help practitioners choose the most appropriate solution for their use case. We primarily evaluated each technology in terms of latency and throughput. Our experiments were conducted using a diverse array of workloads to test the message queues under various scenarios. This enables practitioners to evaluate the performance of the systems and choose the one that best meets their needs. The results show that each technology has its own pros and cons. Specifically, Redis performed the best in terms of latency, whereas Kafka significantly outperformed the other three technologies in terms of throughput. The optimal choice depends on the specific requirements of the use case. This paper presents valuable insights for practitioners and researchers working with message queues. Furthermore, the results of our experiments are provided in JSON format as a supplement to this paper.