Informat RabbitMQ Multi-threaded Consumption Configuration and Implementation Guide

1. Overview

This document provides complete configuration, code implementation, and best practices for RabbitMQ multi-threaded consumption in the Informat system, applicable to scenarios requiring improved message processing throughput. The core goal is to achieve multi-threaded concurrent consumption through SpringBoot container thread pool management while ensuring the security and reliability of message processing.

2. Core Principles

RabbitMQ multi-threaded consumption relies on the SimpleMessageListenerContainer (default listening container) provided by SpringBoot, implementing concurrency through the following core configurations:

• Minimum concurrent consumers (concurrency): Fixed base consumption thread count
• Maximum concurrent consumers (max-concurrency): Maximum thread count for dynamic scaling
• Prefetch message count (prefetch): Controls the maximum number of messages prefetched by each consumer to avoid uneven message distribution
• Manual ACK (acknowledge-mode: manual): Compatible with Informat's custom ACK logic, ensuring message reliability

3. Detailed Configuration

3.1 application.yml Configuration

Complete the RabbitMQ connection information and core multi-threaded consumption parameters as follows:

spring:
  rabbitmq:
    # Basic connection configuration
    port: 5672
    host: 127.0.0.1
    username: admin
    password: password
    virtual-host: /
    nickname: Informat-Next
    # Publisher confirmation configuration (optional, ensures producer-side reliability)
    publisher-confirm-type: correlated
    publisher-returns: true
    template:
      mandatory: true
    # Consumer listener container configuration (multi-threaded core)
    listener:
      simple:
        acknowledge-mode: manual  # Manual ACK, compatible with Informat business logic
        concurrency: 5            # Minimum concurrent consumption thread count (adjust according to CPU cores)
        max-concurrency: 10       # Maximum concurrent consumption thread count (dynamic scaling)
        prefetch: 10              # Number of messages prefetched by each consumer (avoids message accumulation)
        default-requeue-rejected: false  # Failed messages do not return to queue (adjust according to business)
        retry:
          enabled: false  # Disable Spring default retry (Informat business handles failure logic independently)