Quick Operation Guide

# 1. Deployment Architecture

# Overall Architecture

┌─────────────────────────────────────────────────────────────┐
│                        客户端层                               │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────────────┐  │
│  │ 生产者应用   │  │ 消费者应用   │  │ 管控台/Web管理       │  │
│  └─────────────┘  └─────────────┘  └─────────────────────┘  │
└────────────────────────┬────────────────────────────────────┘
                         │ AMQP/MQTT/STOMP
┌────────────────────────▼────────────────────────────────────┐
│                     消息引擎层                                │
│  ┌───────────────────────────────────────────────────────┐  │
│  │              ADMQ for RabbitMQ 节点                    │  │
│  │  ┌─────────────┐  ┌─────────────┐  ┌──────────────┐  │  │
│  │  │  连接层      │  │  交换机      │  │   队列        │  │  │
│  │  │ Connection  │  │  Exchange   │  │   Queue      │  │  │
│  │  └─────────────┘  └─────────────┘  └──────────────┘  │  │
│  │                                                        │  │
│  │  ┌─────────────┐  ┌─────────────┐  ┌──────────────┐  │  │
│  │  │   路由层     │  │  镜像队列    │  │  持久化存储   │  │  │
│  │  │   Router    │  │Mirror Queue │  │  Mnesia/ETS  │  │  │
│  │  └─────────────┘  └─────────────┘  └──────────────┘  │  │
│  └───────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────┘
                         │
┌────────────────────────▼────────────────────────────────────┐
│                      集群通信层                               │
│  ┌───────────────────────────────────────────────────────┐  │
│  │              Erlang Distribution 网络                  │  │
│  │            (节点间元数据同步与心跳检测)                  │  │
│  └───────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────┘
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

# Deployment Mode

Deployment Mode Nodes Use Case High Availability
Standalone 1 node Development/Testing None
Standard Cluster 2+ nodes Production Environment Metadata HA
Mirror Queue Cluster 3+ nodes Production Environment Metadata + Data HA

# 2. Quick Installation

# 2.1 Environment Preparation

Component Minimum Requirement Recommended Configuration
CPU 4 cores 8 cores or more
Memory 8 GB 16 GB or more
Disk 100 GB SSD 500 GB or more SSD
OS CentOS 7+, Ubuntu 18.04+, Kylin, UOS -

Port Requirements:

Port Purpose
5672 AMQP protocol
15672 Web management interface
25672 Cluster inter-node communication

# 2.2 Standalone Installation

# Create directory
mkdir -p /apusic && cd /apusic

# Extract installation package
tar zxvf ADMQ-V2.0.6.xxx.tar.gz
cd admq-rabbitmq

# Place license.xml license file in the root directory
# Modify hostname (only letters, digits, and hyphens allowed)
hostname node1

# Start service
bin/admq-daemon start rabbitmq server
1
2
3
4
5
6
7
8
9
10
11
12
13

# 2.3 Cluster Installation (3 Nodes)

Execute on all nodes:

mkdir -p /apusic && cd /apusic
tar zxvf ADMQ-V2.0.6.xxx.tar.gz
cd admq-rabbitmq

# Place license.xml
# Modify hostname
hostname node171  # node172, node173

# Configure /etc/hosts
172.20.140.171 node171
172.20.140.172 node172
172.20.140.173 node173

# Start service
bin/admq-daemon start rabbitmq server
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

Execute cluster creation on Node 1:

bin/admq rabbitmq cluster add -s 172.20.140.172:58181 -u monitor -p apusic_123
bin/admq rabbitmq cluster add -s 172.20.140.173:58181 -u monitor -p apusic_123
1
2

Enable mirror queue (optional):

bin/admq rabbitmq admin ctl set_policy ha-all "^" '{"ha-mode":"all","ha-sync-mode":"automatic"}'
1

# 3. Core Concepts

# Message Model

Producer -> Exchange -> [Binding/Routing Key] -> Queue -> Consumer
1

# Exchange Types

Type Routing Method Use Case
Direct Exact Routing Key match Point-to-point messaging
Fanout Broadcast to all queues Broadcast notifications
Topic Pattern matching * # Log categorization, event routing
Headers Match based on message headers Complex conditional routing

# Queue Types

Type Characteristics Use Case
Classic Queue Traditional queue, supports mirroring General business scenarios
Quorum Queue Raft-based, automatic multi-replica High availability scenarios
Stream Queue Append-only log, supports multiple reads High throughput, logging scenarios

# 4. Common Command Reference

# Service Management

# Start/Stop/Restart
bin/admq-daemon start rabbitmq server
bin/admq-daemon stop rabbitmq server
bin/admq-daemon restart rabbitmq server

# Application start/stop
bin/admq rabbitmq admin ctl start_app
bin/admq rabbitmq admin ctl stop_app
1
2
3
4
5
6
7
8

# Status Inspection

# Node status
bin/admq rabbitmq admin ctl status

# Cluster status
bin/admq rabbitmq admin ctl cluster_status

# Health check
bin/admq rabbitmq admin diagnostics health_check

# Memory usage
bin/admq rabbitmq admin diagnostics memory_breakdown
1
2
3
4
5
6
7
8
9
10
11

# User Management

# Create user
bin/admq rabbitmq admin ctl add_user username password
bin/admq rabbitmq admin ctl set_user_tags username administrator
bin/admq rabbitmq admin ctl set_permissions -p /vhost username ".*" ".*" ".*"

# Change password
bin/admq rabbitmq admin ctl change_password username newpassword

# Delete user
bin/admq rabbitmq admin ctl delete_user username

# List users
bin/admq rabbitmq admin ctl list_users
1
2
3
4
5
6
7
8
9
10
11
12
13

# Queue Management

# List queues
bin/admq rabbitmq admin ctl list_queues
bin/admq rabbitmq admin ctl list_queues name messages consumers

# List exchanges
bin/admq rabbitmq admin ctl list_exchanges name type

# List connections
bin/admq rabbitmq admin ctl list_connections peer_host state

# List channels
bin/admq rabbitmq admin ctl list_channels connection consumer_count
1
2
3
4
5
6
7
8
9
10
11
12

# Cluster Management

# View cluster status
bin/admq rabbitmq admin ctl cluster_status

# View queue details (including mirror)
bin/admq rabbitmq admin ctl list_queues name pid slave_pids synchronised_slave_pids

# Remove failed node
bin/admq rabbitmq admin ctl forget_cluster_node rabbit@node2
1
2
3
4
5
6
7
8

# Policy Management

# List policies
bin/admq rabbitmq admin ctl list_policies

# Set mirror policy
bin/admq rabbitmq admin ctl set_policy ha-all "^" '{"ha-mode":"all"}' --priority 0 --apply-to queues

# Clear policy
bin/admq rabbitmq admin ctl clear_policy ha-all
1
2
3
4
5
6
7
8

# Backup and Restore

# Export configuration
bin/admq rabbitmq admin ctl export_definitions /backup/definitions.json

# Import configuration
bin/admq rabbitmq admin ctl import_definitions /backup/definitions.json

# Backup Mnesia
bin/admq rabbitmq admin ctl eval 'rabbit_mnesia:backup("/backup/mnesia.backup").'
1
2
3
4
5
6
7
8

# 5. Daily Operations

# Log Locations

Log Type Path
Service log ./log/rabbitmq/rabbit@hostname.log
Startup log ./log/rabbitmq/rabbit@hostname-sasl.log
Console log ./logs/

# Daily Checklist

Check Item Command Normal Standard
Node status bin/admq rabbitmq admin ctl status Running normally
Cluster status bin/admq rabbitmq admin ctl cluster_status All nodes online
Queue backlog bin/admq rabbitmq admin ctl list_queues messages No abnormal backlog
Memory usage bin/admq rabbitmq admin ctl status Below watermark
Disk space df -h Above disk_free_limit

# Health Check Script

#!/bin/bash
# Node health check
bin/admq rabbitmq admin diagnostics health_check

# Memory check
bin/admq rabbitmq admin diagnostics memory_breakdown

# Alert queue check (messages > 10000)
bin/admq rabbitmq admin ctl list_queues name messages consumers | awk '$2 > 10000 {print "WARNING: Queue "$1" has "$2" messages"}'

# Connection count check
CONN_COUNT=$(bin/admq rabbitmq admin ctl list_connections | wc -l)
if [ $CONN_COUNT -gt 1000 ]; then
    echo "WARNING: Connection count $CONN_COUNT exceeds threshold"
fi
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

# 6. Configuration Optimization

# rabbitmq.conf Core Configuration

# Network
listeners.tcp.default = 5672
listeners.ssl.default = 5671

# Memory (recommended 0.6-0.7)
vm_memory_high_watermark.relative = 0.6
vm_memory_high_watermark_paging_ratio = 0.5

# Disk (recommended absolute value 1GB)
disk_free_limit.relative = 1.5
disk_free_limit.absolute = 1GB

# Connections
channel_max = 2047
heartbeat = 30

# Queues
queue_master_locator = min-masters
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

# System Optimization

# File descriptors
ulimit -n 65535

# Kernel parameters (/etc/sysctl.conf)
net.ipv4.ip_local_port_range = 1024 65535
net.ipv4.tcp_tw_reuse = 1
net.core.somaxconn = 65535
vm.swappiness = 10

# Disk scheduler
echo deadline > /sys/block/sda/queue/scheduler
1
2
3
4
5
6
7
8
9
10
11

# 7. Troubleshooting

# Service Won't Start

Symptom Possible Cause Solution
Port occupied 5672/15672 occupied Stop the occupying process or change the port
Cookie mismatch Cluster node Cookie mismatch Synchronize .erlang.cookie
Insufficient memory System memory insufficient Increase memory or lower watermark
Insufficient disk Below disk_free_limit Clean up disk or adjust limit

# Message Consumption Anomalies

Symptom Possible Cause Solution
Message backlog Consumer too slow or offline Add more consumers or optimize logic
Message loss Persistence not enabled Enable delivery_mode=2
Duplicate consumption Acknowledgement timeout Optimize processing logic, acknowledge promptly

# Cluster Anomalies

Symptom Possible Cause Solution
Node offline Network failure Check network, restart service
Split-brain Network partition Use forget_cluster_node to clean up
Queue unavailable Node offline Use mirror queue or Quorum Queue

# Diagnostic Commands

# View detailed information
bin/admq rabbitmq admin ctl status
bin/admq rabbitmq admin ctl cluster_status
bin/admq rabbitmq admin ctl list_connections peer_host peer_port state
bin/admq rabbitmq admin ctl list_queues name messages consumers memory

# View memory/processes
bin/admq rabbitmq admin diagnostics memory_breakdown
bin/admq rabbitmq admin diagnostics processes
bin/admq rabbitmq admin diagnostics connections
1
2
3
4
5
6
7
8
9
10

# 8. Security Hardening

# Change Default User

# Delete default guest user
bin/admq rabbitmq admin ctl delete_user guest

# Create new administrator
bin/admq rabbitmq admin ctl add_user admin strongpassword
bin/admq rabbitmq admin ctl set_user_tags admin administrator
bin/admq rabbitmq admin ctl set_permissions -p / admin ".*" ".*" ".*"
1
2
3
4
5
6
7

# Firewall Configuration

firewall-cmd --permanent --add-port=5672/tcp
firewall-cmd --permanent --add-port=15672/tcp
firewall-cmd --permanent --add-port=25672/tcp
firewall-cmd --reload
1
2
3
4

# SSL Configuration

listeners.ssl.default = 5671
ssl_options.cacertfile = /path/to/ca_certificate.pem
ssl_options.certfile = /path/to/server_certificate.pem
ssl_options.keyfile = /path/to/server_key.pem
ssl_options.verify = verify_peer
ssl_options.fail_if_no_peer_cert = false
1
2
3
4
5
6

# 9. Performance Optimization Tips

# Queue Optimization

Map<String, Object> args = new HashMap<>();
// Queue length limit
args.put("x-max-length", 100000);
args.put("x-max-length-bytes", 1073741824);  // 1GB

// Message TTL
args.put("x-message-ttl", 3600000);  // 1 hour

// Lazy queue (reduce memory)
args.put("x-queue-mode", "lazy");
1
2
3
4
5
6
7
8
9
10

# Client Optimization

// Connection pool + automatic recovery
factory.setAutomaticRecoveryEnabled(true);
factory.setNetworkRecoveryInterval(5000);

// Single connection with multiple channels
Connection connection = factory.newConnection();
for (int i = 0; i < 10; i++) {
    Channel channel = connection.createChannel();
}

// QoS prefetch
channel.basicQos(100);

// Batch acknowledgment
channel.basicAck(deliveryTag, true);
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

# Monitoring Metrics

Metric Description
rabbitmq_connections Number of connections
rabbitmq_queue_messages Total message count
rabbitmq_node_mem_used Memory usage
rabbitmq_node_disk_free Available disk space

# 10. Operations Best Practices

  1. Regular Checks: Establish routine inspection mechanisms
  2. Monitoring and Alerts: Configure a comprehensive monitoring and alerting system
  3. Backup Strategy: Regularly back up configurations and data
  4. Capacity Planning: Evaluate and scale up in advance
  5. Security Updates: Apply security patches in a timely manner
  6. Least Privilege: Assign minimum necessary permissions to business users
  7. Disaster Drills: Conduct regular failure recovery drills

Technical Support:

  • Website: www.apusic.com
  • Phone: 400-855-5800
  • Email: support@apusic.com
编辑页面 (opens new window)
#运维手册

← 產品介紹→