Why RocketMQ’s Ordinary Messages Power Business Integration

Business Integration vs. Data Integration

Business integration focuses on real‑time order flow, millisecond‑level latency, and tightly coupled service interactions. Data integration emphasizes aggregation and analysis of heterogeneous data sources, tolerating near‑real‑time or batch processing.

Core Messaging Requirements for Business Integration

Support for multiple message types: ordinary, scheduled, transactional, ordered.

Rich routing: tag filtering, attribute filtering, one‑to‑many and one‑to‑one dispatch.

Various interaction modes: synchronous, asynchronous, push‑pull, streaming responses.

Observability: metrics, tracing, events, full‑link tracing, alerts.

Ordinary Message Fundamentals

Ordinary messages are the simplest and most widely used RocketMQ message type, offering high throughput, horizontal scalability, low latency, and asynchronous communication.

Key Characteristics

Atomic processing – each message is independent.

Scalable – multiple queues, horizontal sharding, concurrent consumption.

Low latency – short processing path enables millisecond‑level delivery.

Message Lifecycle

Initialization : Producer constructs the message (topic, tag, key, payload) and sends it.

Pending consumption : Broker stores the message and makes it visible to consumers.

Consuming : Consumer fetches the message; if no ack is received, the broker retries.

Consume‑ack : Consumer acknowledges success or failure; the message remains retained for possible replay.

Deletion : After the configured retention period, the broker physically removes the message.

Typical Use Cases

Microservice decoupling – asynchronous calls reduce response time and smooth traffic spikes.

Real‑time data transmission – high‑throughput pipelines for telemetry, logs, or event streams.

Case Study: E‑commerce Transaction Platform

During large‑scale sales events, RocketMQ ordinary messages decouple order processing, payment, logistics, and inventory systems. The architecture must handle complex order state machines, massive order spikes, and ensure transactional consistency across services.

Practical Code Walkthrough

Sending an Ordinary Message

Key steps:

Fully initialize the message with topic, tag, key, and payload.

Capture the send result and handle exceptions for retry logic.

Message msg = new Message("TopicTest", "TagA", "Key123", "Hello RocketMQ".getBytes());
SendResult result = producer.send(msg);
System.out.printf("Send status: %s, msgId: %s%n", result.getSendStatus(), result.getMsgId());

Consuming an Ordinary Message

RocketMQ provides two consumption models:

Passive listener – register a MessageListenerConcurrently (or Orderly) and return ConsumeConcurrentlyStatus.CONSUME_SUCCESS on success; throw an exception or return RECONSUME_LATER to trigger retry.

Active pull – use DefaultMQPullConsumer to pull messages on demand, allowing custom rate‑limiting and concurrency control.

DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("ConsumerGroupA");
consumer.subscribe("TopicTest", "*");
consumer.registerMessageListener((msgs, context) -> {
    for (MessageExt msg : msgs) {
        System.out.println(new String(msg.getBody()));
    }
    return ConsumeConcurrentlyStatus.CONSUME_SUCCESS;
});
consumer.start();

Observability Features

RocketMQ integrates metrics, tracing, and event reporting. It supports single‑link and full‑link traceability, enabling monitoring dashboards and alerting for latency, throughput, and error rates.

Future Outlook

Alibaba Cloud plans to release RocketMQ 5.0, which will provide higher elasticity, lower cost, and simplified operations, further strengthening its suitability for mission‑critical business integration scenarios.