Comprehensive Overview of Message Queues: Types, Core Concepts, and Comparison of Kafka, RocketMQ, and RabbitMQ

1. Full List of Message Queues

Commonly used message queue engines include ZeroMQ, DistributedLog, ActiveMQ, RabbitMQ, Kafka, RocketMQ, Artemis, Apollo, IronMQ, OpenMQ, and Pulsar, covering both open‑source and commercial options.

2. Technical Applications of MQ

Decoupling: MQs separate producers from consumers, reducing integration complexity.

Eventual Consistency: MQs can help achieve state consistency across systems, ensuring either all succeed or all fail.

Broadcast: MQs enable one‑to‑many message distribution, simplifying subscriber management.

Throttling & Flow Control: MQs smooth traffic spikes, a typical use case being flash‑sale (秒杀) traffic shaping.

3. Comparison of Major MQs

ActiveMQ

Pros: High availability via master‑slave, millisecond latency, rich feature set, and decent throughput (tens of thousands TPS).

Cons: Declining community support and limited suitability for massive scale.

Kafka

Pros: Extremely high throughput (millions TPS), millisecond latency, strong durability with replication, pull‑based consumption, mature ecosystem (e.g., Kafka‑Manager), and wide adoption in big‑data pipelines.

Cons: Performance degrades with many partitions, reliance on short‑polling, limited retry on consumer failures, possible ordering issues after broker loss, and slower release cycle.

RabbitMQ

Pros: Good performance due to Erlang, high concurrency, robust feature set, cross‑platform support, excellent management UI, and active community.

Cons: Lower throughput compared to Kafka, heavier implementation, steep learning curve for protocols, and limited source‑code readability.

RocketMQ

Pros: High single‑node throughput (hundreds of thousands TPS), strong availability, near‑zero message loss with proper tuning, rich feature set, support for massive message backlogs, and Java source code for easy customization.

Cons: Fewer client language SDKs, moderate community activity, and lack of native JMS support requiring code changes for migration.

Choosing the right MQ depends on the scenario: RabbitMQ for general use, Kafka for big‑data real‑time processing, and RocketMQ for high‑transactional e‑commerce or financial systems.