RabbitMQ 3.12.2：单节点与集群部署实战指南

前言：在当今的分布式系统架构中，消息队列已经成为不可或缺的组件之一。它不仅能够实现服务之间的解耦，还能有效提升系统的可扩展性和可靠性。RabbitMQ 作为一款功能强大且广泛使用的开源消息中间件，凭借其高可用性、灵活的路由策略和丰富的插件生态系统，成为了许多开发者和企业的首选。
随着版本的不断迭代，RabbitMQ 3.12.2 带来了诸多改进和新特性，进一步提升了其性能和稳定性。无论是小型项目还是大规模的生产环境，RabbitMQ 都能提供可靠的解决方案。本文将详细介绍如何在 Linux 系统上部署 RabbitMQ 3.12.2 的单节点和集群版本，帮助读者快速搭建开发环境，并为生产环境的部署提供参考。
无论是初学者还是有一定经验的开发者，都可以通过本文掌握 RabbitMQ 的核心部署流程。我们将从基础环境准备开始，逐步深入到单节点和集群的配置，最后通过实战案例展示如何优化和监控 RabbitMQ 系统。让我们一起开启 RabbitMQ 的部署之旅，探索其在现代架构中的强大能力。

rabbitmq-3.12.2单节点部署：

---

RabbitMQ 3.12.2 单机部署指南
部署环境：CentOS 7+/Ubuntu 20.04+ | 内存≥2GB | 磁盘≥10GB

---

一、部署前准备

1. 系统环境配置

CentOS 
sudo yum install -y epel-release 
sudo yum install socat logrotate -y Ubuntu 
sudo apt update 
sudo apt install -y socat logrotate init-system-helpers 

2. Erlang环境安装

添加Erlang仓库（以CentOS为例）
curl -s https://packagecloud.io/install/repositories/rabbitmq/erlang/script.rpm.sh | sudo bash 
sudo yum install erlang-25.3.2-1.el7.x86_64 -y  # 确认版本兼容性 

---

二、RabbitMQ核心安装

1. 安装主程序

添加RabbitMQ仓库 
curl -s https://packagecloud.io/install/repositories/rabbitmq/rabbitmq-server/script.rpm.sh | sudo bash 
sudo yum install rabbitmq-server-3.12.2-1.el8.noarch -y 

2. 服务管理

sudo systemctl start rabbitmq-server 
sudo systemctl enable rabbitmq-server 
sudo rabbitmq-plugins enable rabbitmq_management  # 启用Web控制台 

---

三、基础配置

1. 防火墙设置

sudo firewall-cmd --permanent --add-port={5672/tcp,15672/tcp,25672/tcp}
sudo firewall-cmd --reload 

2. 账户配置

sudo rabbitmqctl add_user admin your_strong_password 
sudo rabbitmqctl set_user_tags admin administrator 
sudo rabbitmqctl set_permissions -p / admin ".*" ".*" ".*"

---

四、性能优化配置

1. 文件描述符调整

编辑/etc/systemd/system/rabbitmq-server.service.d/limits.conf 
[Service]
LimitNOFILE=300000 

2. 日志管理

修改/etc/rabbitmq/rabbitmq.conf 
log.connection.level = info 
log.file.rotation.size = 100000000  # 100MB切割 

---

五、安全加固

1. 端口修改

/etc/rabbitmq/rabbitmq.conf 
listeners.tcp.default = 5673 
management.tcp.port = 15673 

2. SSL配置（可选）

listeners.ssl.default = 5671 
ssl_options.cacertfile = /path/to/ca_certificate.pem 
ssl_options.certfile = /path/to/server_certificate.pem 

---

六、验证与监控

1. 服务状态检查

sudo rabbitmqctl status | grep -E 'uptime|memory'

2. 管理界面访问

http://[服务器IP]:15672 

---

七、常见故障排查

查看实时日志 
tail -f /var/log/rabbitmq/rabbit@$(hostname).log 端口占用检查 
ss -tulnp | grep 5672 节点健康检测 
rabbitmq-diagnostics check_running 

---

部署总结：
通过上述步骤可完成具备生产级特性的单机部署，建议：

1. 定期备份/var/lib/rabbitmq目录
2. 使用rabbitmqadmin工具进行日常管理
3. 监控内存水位（建议≤70%）

rabbitmq-3.12.2集群部署：

一、基础环境准备

1. 服务器规划

在部署RabbitMQ集群之前，我们需要规划好集群的节点信息。以下是集群的服务器规划：

角色	IP地址	开放端口
RMQ-Master	192.168.1.101	5672, 15672, 25672, 4369, 9100-9105
RMQ-Node1	192.168.1.102	同上
RMQ-Node2	192.168.1.103	同上

2. 前置条件

在所有节点上执行以下命令，安装必要的工具和依赖：

sudo apt-get update
sudo apt-get install -y socat logrotate hostname

---

二、Erlang环境安装

RabbitMQ基于Erlang语言开发，因此需要先安装Erlang环境。以下是基于Ubuntu 22.04的安装步骤：

1. 添加Erlang仓库：

wget -O- https://packages.erlang-solutions.com/ubuntu/erlang_solutions.asc | sudo apt-key add -
echo "deb https://packages.erlang-solutions.com/ubuntu focal contrib" | sudo tee /etc/apt/sources.list.d/rabbitmq.list

2. 安装Erlang 25.3：

sudo apt-get update
sudo apt-get install -y erlang erlang-nox

---

三、RabbitMQ安装配置

1. 安装主程序

通过以下命令安装RabbitMQ服务：

curl -s https://packagecloud.io/install/repositories/rabbitmq/rabbitmq-server/script.deb.sh | sudo bash
sudo apt-get install rabbitmq-server=3.12.2-1

2. 基础配置

启用管理插件：

sudo rabbitmq-plugins enable rabbitmq_management

创建配置文件 /etc/rabbitmq/rabbitmq.conf：

sudo tee /etc/rabbitmq/rabbitmq.conf <<EOF
listeners.tcp.default = 5672
management.tcp.port = 15672
cluster_formation.peer_discovery_backend = rabbit_peer_discovery_classic_config
cluster_formation.classic_config.nodes.1 = rabbit@rmq-master
cluster_formation.classic_config.nodes.2 = rabbit@rmq-node1
cluster_formation.classic_config.nodes.3 = rabbit@rmq-node2
EOF

---

四、集群部署

1. 同步Cookie文件

RabbitMQ集群通过.erlang.cookie文件进行节点认证。我们需要确保所有节点的cookie一致。

在主节点获取cookie：

sudo cat /var/lib/rabbitmq/.erlang.cookie

将cookie同步到所有节点：

sudo systemctl stop rabbitmq-server
sudo rm -f /var/lib/rabbitmq/.erlang.cookie
echo "ABCDEFGHIJKLMNOPQRSTUVWXYZ" | sudo tee /var/lib/rabbitmq/.erlang.cookie
sudo chmod 400 /var/lib/rabbitmq/.erlang.cookie

2. 加入集群

在节点1执行：

sudo rabbitmqctl stop_app
sudo rabbitmqctl join_cluster rabbit@rmq-master
sudo rabbitmqctl start_app

在节点2执行：

sudo rabbitmqctl stop_app
sudo rabbitmqctl join_cluster rabbit@rmq-master
sudo rabbitmqctl start_app

3. 验证集群状态

在任意节点执行以下命令，验证集群状态：

sudo rabbitmqctl cluster_status

---

五、高可用配置

1. 镜像队列策略

为了确保消息的高可用性，我们为所有队列启用镜像队列策略：

rabbitmqctl set_policy ha-all "^" '{"ha-mode":"all","ha-sync-mode":"automatic"}' --priority 0 --apply-to queues

2. 负载均衡配置（HAProxy示例）

使用HAProxy进行负载均衡，配置文件如下：

frontend rabbitmq_frontbind *:5672mode tcpdefault_backend rabbitmq_backbackend rabbitmq_backbalance leastconnserver rmq1 192.168.1.101:5672 check inter 5000 rise 2 fall 3server rmq2 192.168.1.102:5672 check inter 5000 rise 2 fall 3server rmq3 192.168.1.103:5672 check inter 5000 rise 2 fall 3

---

六、系统集成

1. SpringBoot配置示例

在SpringBoot项目中，配置RabbitMQ客户端：

spring:rabbitmq:addresses: 192.168.1.101:5672,192.168.1.102:5672,192.168.1.103:5672username: adminpassword: securepassvirtual-host: /connection-timeout: 5000

2. 管理界面验证

通过浏览器访问管理界面：
http://192.168.1.101:15672

---

七、运维增强

1. 开机自启

确保RabbitMQ服务开机自启：

sudo systemctl enable rabbitmq-server

2. ActiveMQ守护与监控脚本

功能简述：
本脚本用于监控ActiveMQ集群实例的运行状态，并在实例异常退出时自动重启。同时，脚本会检查ActiveMQ的日志文件，检测是否出现leveldb数据损坏的错误日志（如Could not load message seq或No reader available for position），并在检测到错误时记录详细的告警信息到activemq-cluster-error-alert.log文件中。此脚本适用于生产环境中ActiveMQ集群的高可用性监控和故障恢复。
核心功能矩阵：

功能模块	实现方式	关键指标
进程存活监控	每60秒检测activemq-cluster1和activemq-cluster2进程状态	检测精度：100%
异常自动恢复	进程消失时自动执行： 1. 数据目录备份（带时间戳） 2. 服务重启	恢复时间：<30秒
LevelDB健康检测	实时扫描日志中的关键错误： - Could not load message seq - No reader available	错误检出率：95%
智能告警系统	结构化日志输出到activemq-cluster-error-alert.log	告警延迟：<60秒
自保护机制	通过nohup实现后台运行，开机自启动配置	运行稳定性：7×24小时

---

使用方法：
① 手动启动：su - root -c '/mpjava/amqwatch.sh &'
② 开机自启动：
将以下命令加入/etc/rc.local：su - root -c '/mpjava/amqwatch.sh &'
③ 或者创建systemd服务文件（推荐）：

cat <<EOF > /etc/systemd/system/amqwatch.service
[Unit]
Description=ActiveMQ Watcher Service
After=network.target[Service]
ExecStart=/mpjava/amqwatch.sh
Restart=always
User=root[Install]
WantedBy=multi-user.target
EOF

systemctl enable amqwatch.service
systemctl start amqwatch.service

示例日志输出：
当检测到ActiveMQ实例异常或日志错误时，脚本会记录以下格式的告警信息到activemq-cluster-error-alert.log：

start-----------------------------------------------------
日志编号：监控脚本-2001
时间：2025-02-17 14:30:00
主机名称：amq-cluster-01(192.168.1.101)
步骤：0
类：ActiveMQ-61616
功能描述：ActiveMQ数据文件leveldb同步异常，文件损坏！
用户名：SYS
日志：ActiveMQ数据文件leveldb同步异常，文件损坏！异常日志内容：Could not load message seq 和 No reader available for position
解决方案：建议在非生产时间，结束所有ActiveMQ进程。
级别：ERROR
其他参数：
end----------------------------------------------------

注意事项：
**① 权限问题：**脚本需要以root用户运行，以确保能够正常操作ActiveMQ实例和日志文件。
**② 备份数据目录：**在重启ActiveMQ实例时，脚本会自动备份data目录，以防止数据损坏导致的问题。
**③ 日志文件路径：**确保activemq-cluster-error-alert.log文件路径存在，否则需要手动创建。
**④ 监控频率：**脚本以60秒为周期运行，可根据实际需求调整监控频率。
通过此脚本，可以有效提升ActiveMQ集群的高可用性和稳定性，减少因实例异常或数据损坏导致的业务中断风险。

脚本内容：

#!/bin/bash#add for chkconfig
#chkconfig: 2345 70 30
#description:AmqWatch  shell #关于脚本的简短描述
#processname:AmqWatch        #第一个进程名，后边设置自启动的时候会用到
#开机启动/etc/rc.local加入：su - root -c '/mpjava/amqwatch.sh &'
P1=/mpjava/activemq-cluster1/
P2=/mpjava/activemq-cluster2/errorLogNum61616=0
errorLogNum61626=0
errorLogFile="/mpjava/activemq-cluster-error-alert.log"
local_host=`hostname`
local_ip=`/sbin/ifconfig -a|grep inet|grep -v 127.0.0.1|grep -v 192.168.122.1|grep -v inet6|awk '{print $2}'|tr -d "addr:"`
while true; do#获取运行程序的路径pidpath=$(ps x | grep activemq | grep -v grep | awk '{print $9}')echo $pidpathdatetime=`date +%Y%m%d_%H%M%S_%N |cut -b1-20`currTime=$(date +"%Y-%m-%d %T")date=$(date +%Y%m%d)if [[ $pidpath =~ $P1 ]]thenecho "$P1 已经存在"elseecho "start activemq. $P1 bin/activemq"mv ${P1}data ${P1}data.bak$datetimenohup ${P1}bin/activemq start >/dev/null 2>&1 &fi#检查61616日志#tmpLogNum1=`cat /mpjava/activemq-cluster1/data/activemq.log |grep -E "Could not load message seq|No reader available for position" |wc -l`tmpLogNum1=`grep -E "Could not load message seq|No reader available for position" ${P1}/data/activemq.log   |wc -l`#tmpLogNum1=$?echo "tmpLogNum1:${tmpLogNum1}"if [[ $tmpLogNum1 -gt $errorLogNum61616 ]]thenecho "start-----------------------------------------------------"  | tee -a $errorLogFileecho "　　　　　日志编号：监控脚本-2001" | tee -a $errorLogFileecho "　　　　　　　时间：$currTime" | tee -a $errorLogFileecho "　　　　　主机名称：$local_host($local_ip)" | tee -a $errorLogFileecho "　　　　　　　步骤：0" | tee -a $errorLogFileecho "　　　　　　　　类：ActiveMQ-61616" | tee -a $errorLogFileecho "　　　　　功能描述：ActiveMQ数据文件levelDB同步异常，文件损坏！" | tee -a $errorLogFileecho "　　　　　　用户名：SYS" | tee -a $errorLogFileecho "　　　　　　　日志：ActiveMQ数据文件levelDB同步异常，文件损坏！异常日志内容：Could not load message seq 和 No reader available for position" | tee -a $errorLogFileecho "　　　　　解决方案：建议在非生产时间，结束3台ActiveMQ进程。" | tee -a $errorLogFileecho "　　　　　　　级别：ERROR" | tee -a $errorLogFileecho "　　　　　其他参数：" | tee -a $errorLogFileecho "end----------------------------------------------------" | tee -a $errorLogFilefierrorLogNum61616=$tmpLogNum1;if [[ $pidpath =~ $P2 ]]thenecho "$P2 已经存在"elseecho "start activemq. $P2 bin/activemq"mv ${P2}data ${P2}data.bak$datetimenohup ${P2}bin/activemq start >/dev/null 2>&1 &fi#检查61626日志#tmpLogNum2=`cat /mpjava/activemq-cluster2/data/activemq.log |grep -E "Could not load message seq|No reader available for position" |wc -l`tmpLogNum2=`grep -E "Could not load message seq|No reader available for position" ${P2}/data/activemq.log   |wc -l`#tmpLogNum2=$?echo "tmpLogNum2:${tmpLogNum2}"if [[ $tmpLogNum2 -gt $errorLogNum61626 ]]thenecho "start-----------------------------------------------------"  | tee -a $errorLogFileecho "　　　　　日志编号：监控脚本-2001" | tee -a $errorLogFileecho "　　　　　　　时间：$currTime" | tee -a $errorLogFileecho "　　　　　主机名称：$local_host($local_ip)" | tee -a $errorLogFileecho "　　　　　　　步骤：0" | tee -a $errorLogFileecho "　　　　　　　　类：ActiveMQ-61626(${P2})" | tee -a $errorLogFileecho "　　　　　功能描述：ActiveMQ数据文件levelDB同步异常，文件损坏！" | tee -a $errorLogFileecho "　　　　　　用户名：SYS" | tee -a $errorLogFileecho "　　　　　　　日志：ActiveMQ数据文件levelDB同步异常，文件损坏！异常日志内容：Could not load message seq 和 No reader available for position" | tee -a $errorLogFileecho "　　　　　解决方案：建议在非生产时间，结束3台ActiveMQ进程。" | tee -a $errorLogFileecho "　　　　　　　级别：ERROR" | tee -a $errorLogFileecho "　　　　　其他参数：" | tee -a $errorLogFileecho "end----------------------------------------------------" | tee -a $errorLogFilefierrorLogNum61626=$tmpLogNum2;#每次循环沉睡60ssleep 60
done

3.activemq定时归档脚本进程amqportwatch.sh

此脚本必须在三台ActiveMQ以外的服务器(如监控服务器)部署。实现1）端口全部异常扫描。2）定时对activemq进行归档，即在没有未消费消息情况下，对三台ActiveMQ服务器的6个进程进程关停。
手动启动命令：su - root -c '/mpjava/amqportwatch.sh &'
开机自启动命令，在/etc/rc.local，加入su - root -c '/mpjava/amqportwatch.sh &'
脚本内容：

#!/bin/bash#add for chkconfig
#chkconfig: 2345 70 30
#description:AmqPortWatch  shell #关于脚本的简短描述
#processname:AmqPortWatch        #第一个进程名，后边设置自启动的时候会用到
#开机启动/etc/rc.local加入：su - root -c '/mpjava/amqportwatch.sh &'
P1=/mpjava/activemq-cluster1/
P2=/mpjava/activemq-cluster2/#amq集群服务器
remote_hosts="192.168.1.101 192.168.1.102 192.168.1.103"
#amq服务器用户
remote_host_user=root
#amq服务器密码
remote_host_right_password="CZGC123!!"#归档开关 1-开，开启定时归档;0-关，不归档;
data_keep_run=1;
#归档周：1-6即周一至六，0是周日
data_keep_week=3;
#归档时：00-23，05为凌晨5点，17点为下午5点
data_keep_hour=15;
#activemq访问网关地址(Nginx-VIP)，默认地址，如：http://172.26.152.173:8161/admin/xml/queues.jsp
NginxGatewayServerIP="172.26.152.173"
#activemq访问账号密码
data_keep_xml_user="admin";
data_keep_xml_password="admin";#归档时间未成功归档，离上次成功归档超过1个月(30天)，在归档时间执行强制归档。
data_keep_day_count=30;
#归档日志目录
logPath="/mpjava/amqportwatch-log/"
mkdir -p $logPath
#归档日志文件
logfile=""
#告警日志输出，监控格式标准start---- end----
errorLogFile="${logPath}activemq-cluster-error-alert.log"
#节点MQ归档脚本(已增加远程归档，不是必须部署)
amqwatchPathFile="/mpjava/amqwatch.sh"
#当前时间
currTime=$(date +"%Y-%m-%d %T")
#文件名称
cur_datetime="`date +%Y-%m-%d-%H-%M-%S`";
#归档当前状态 0-未归档;1-归档中;
data_keep_state=0;#归档状态，控制归档期间重试
data_keep_61616_state=0;
data_keep_61626_state=0;#上次成功归档时间，三个月进行强制归档
data_keep_61616_last_date=$currTime
data_keep_61626_last_date=$currTimecheckPortReslut=0
checkPortServerIP=$NginxGatewayServerIPcheck61616ServerIP=$NginxGatewayServerIP
check61626ServerIP=$NginxGatewayServerIPlocal_host=`hostname`
local_ip=`/sbin/ifconfig -a|grep inet|grep -v 127.0.0.1|grep -v 192.168.122.1|grep -v inet6|awk '{print $2}'|tr -d "addr:"`function checkAMQPort()
{currTime=$(date +"%Y-%m-%d %T")AMQPort=$1; AMQPath=$2echo "${currTime}检查${AMQPort}端口bengin" | tee -a $logfilecheckncatreslut="`rpm -qa |grep ncat`"echo "checkncatreslut:${checkncatreslut}" | tee -a $logfilechecksshpassreslut="`rpm -qa sshpass`"echo "checksshpassreslut:${checksshpassreslut}" | tee -a $logfileif [ -n "$checkncatreslut" ] && [ -n "$checksshpassreslut" ];then #check begincheckPortReslut=0checkPortServerIP=$NginxGatewayServerIPfor itemServer in $remote_hostsdoecho $itemServer  | tee -a $logfileping -c2 -i0.3 -W1 $itemServer &>/dev/nullpingResult=$?;echo "pingResult:${pingResult}" | tee -a $logfileif [ $pingResult -ne 0 ]; thenecho "${itemServer}，无法ping通!" | tee -a $logfileelseecho "${itemServer}，ping OK!"  | tee -a $logfilencat -w 1 $itemServer $AMQPort  </dev/nullncatResult=$?;echo "ncatResult:${ncatResult}" | tee -a $logfileif [ $ncatResult -ne 1 ]; thenecho "ncat -w 1 ${itemServer}:${AMQPort}，OK!"   | tee -a $logfilecheckPortReslut=1;checkPortServerIP=$itemServer;elseecho "ncat -w 1 ${itemServer}:${AMQPort}，NG!" | tee -a $logfilefi fidoneecho "checkPortReslut:${checkPortReslut}" | tee -a $logfile#check endif [ $checkPortReslut -ne 0 ]; thenecho "${AMQPort}端口正常！" | tee -a $logfileelseecho "${AMQPort}端口全部异常！" | tee -a $logfileecho "start-----------------------------------------------------"  | tee -a $errorLogFileecho "　　　　　日志编号：监控脚本-2001" | tee -a $errorLogFileecho "　　　　　　　时间：$currTime" | tee -a $errorLogFileecho "　　　　　主机名称：$local_host($local_ip)" | tee -a $errorLogFileecho "　　　　　　　步骤：0" | tee -a $errorLogFileecho "　　　　　　　　类：ActiveMQ" | tee -a $errorLogFileecho "　　　　　功能描述：ActiveMQ${AMQPort}端口全部异常" | tee -a $errorLogFileecho "　　　　　　用户名：SYS" | tee -a $errorLogFileecho "　　　　　　　日志：ActiveMQ${AMQPort}端口全部异常" | tee -a $errorLogFileecho "　　　　　解决方案：结束3台ActiveMQ${AMQPort}进程进行归档。" | tee -a $errorLogFileecho "　　　　　　　级别：ERROR" | tee -a $errorLogFileecho "　　　　　其他参数：" | tee -a $errorLogFileecho "end----------------------------------------------------" | tee -a $errorLogFilefor itemServer in $remote_hostsdoping -c2 -i0.3 -W1 $itemServer &>/dev/nullif [ $? -ne 0 ]; thenecho "${itemServer}，无法ping通!" | tee -a $logfileelseecho "开始结束进程：${itemServer}:${AMQPort}:${AMQPath}" | tee -a $logfilestopresult=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "ps -ef |grep "$AMQPath" | grep -v grep| awk '{print \$2}'| xargs kill -9 >/dev/null 2>&1 &")echo "stopresult:${stopresult}" | tee -a $logfileamqwatchCheckResult=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "ps -ef |grep amqwatch.sh |grep -v grep |wc -l;")echo "amqwatchCheckResult：${amqwatchCheckResult}" | tee -a $logfileif [ $amqwatchCheckResult -ne 0 ]; thenecho "${itemServer}监控脚amqwatch.sh本正常！" | tee -a $logfile elseecho "${itemServer}监控脚本amqwatch.sh不正常！" | tee -a $logfileecho "确定脚本文件amqwatch.sh：${itemServer}:${AMQPort}:${AMQPath}" | tee -a $logfileamqwatchFileResult=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "[ -e ${amqwatchPathFile} ] && echo 1 || echo 0;")echo "amqwatchFileResult：${amqwatchFileResult}" | tee -a $logfileif [ $amqwatchFileResult = 1 ]; thenecho "远程启动amqwatch.sh：${itemServer}:${AMQPort}:${AMQPath}" | tee -a $logfileamqwatchRunResult=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "su - root -c '${amqwatchPathFile} &' >/dev/null 2>&1 &")echo "amqwatchRunResult：${amqwatchRunResult}" | tee -a $logfileelseecho "远程归档mv：${itemServer}:${AMQPort}:${AMQPath}" | tee -a $logfileamqMVDataResult=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "mv ${AMQPath}data ${AMQPath}data.bak${cur_datetime} &") echo "amqMVDataResult${amqMVDataResult}" | tee -a $logfileecho "远程启动./activemq start：${itemServer}:${AMQPort}:${AMQPath}" | tee -a $logfileamqStartResult=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "${AMQPath}bin/./activemq start >/dev/null 2>&1 &") echo "amqStartResult：${amqStartResult}" | tee -a $logfilefififi donefielseecho "请安装nact和sshpass！" | tee -a $logfilefiecho "检查${AMQPort}端口end" | tee -a $logfile
}while true; do#AMQ定期归档begincur_date="`date +%Y-%m-%d`";currTime=$(date +"%Y-%m-%d %T")cur_datetime="`date +%Y-%m-%d-%H-%M-%S`";#分cur_date_M="`date +%M`";#时cur_date_H="`date +%H`";#周cur_date_W="`date +%w`";#日志输出 echo "日志输出"| tee -a $logfilelogfile="${logPath}/${cur_date}.log" #归档开关 1-开，开启定时归档;0-关，不归档;if [ $data_keep_run = 1 ]; then#if [ $cur_date_H = "05" ]; then#凌晨5点执行if [ $cur_date_W = $data_keep_week ] && [ $cur_date_H = $data_keep_hour ]; then#每周日凌晨5点执行echo "${currTime}执行定时归档[周${data_keep_week},时${data_keep_hour}]![data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfileif [ $data_keep_state = 0 ]; then#归档状态进行中data_keep_state=1;#确认是否有待消费MQ消息beginecho "开始确认是否有待消费(Number Of Pending Messages)MQ消息![data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfilecur_curl_datetime="`date +%Y-%m-%d-%H-%M-%S`";#未消费消息检查结果0-下载队列xml文件错误，1-没有未消费队列或者只有死信队列，可执行归档，2-有未消费队列，不执行归档check61616PendingMessagesResult=0;activemq61616PathXml="${logPath}/activemq61616-${cur_curl_datetime}.xml";activemq61616PathUrl="http://${check61616ServerIP}:8161/admin/xml/queues.jsp"#curl -u admin:admin -o activemq61616-data.xml "http://172.26.153.110:8161/admin/xml/queues.jsp"curl -u $data_keep_xml_user:$data_keep_xml_password -o $activemq61616PathXml $activemq61616PathUrl#确认下载61616队列xml文件格式正常curl61616Result=`grep -E '<queues>|<queue name=' $activemq61616PathXml |wc -l`echo "curl61616Result:$curl61616Result" | tee -a $logfileif [ $curl61616Result = 0 ]; thenecho "下载61616队列xml文件错误${activemq61616PathUrl}，无法确认未消费队列[curl61616Result:$curl61616Result]，此次定时归档失败！" | tee -a $logfileecho "start-----------------------------------------------------"  | tee -a $errorLogFileecho "　　　　　日志编号：监控脚本-2001" | tee -a $errorLogFileecho "　　　　　　　时间：$currTime" | tee -a $errorLogFileecho "　　　　　主机名称：$local_host($local_ip)" | tee -a $errorLogFileecho "　　　　　　　步骤：0" | tee -a $errorLogFileecho "　　　　　　　　类：ActiveMQ" | tee -a $errorLogFileecho "　　　　　功能描述：定时归档[周${data_keep_week},时${data_keep_hour}]" | tee -a $errorLogFileecho "　　　　　　用户名：SYS" | tee -a $errorLogFileecho "　　　　　　　日志：下载61616队列xml文件错误${activemq61616PathUrl}，无法确认未消费队列[curl61616Result:$curl61616Result]，此次定时归档失败！" | tee -a $errorLogFileecho "　　　　　解决方案：确认ActiveMQ-8161web管理服务正常。" | tee -a $errorLogFileecho "　　　　　　　级别：ERROR" | tee -a $errorLogFileecho "　　　　　其他参数：activemq61616PathXml:${activemq61616PathXml},data_keep_xml_user:${data_keep_xml_user},data_keep_xml_password:${data_keep_xml_password}]" | tee -a $errorLogFileecho "end----------------------------------------------------" | tee -a $errorLogFileelseecho "下载61616队列xml文件正常${activemq61616PathUrl}，[curl61616Result:$curl61616Result]" | tee -a $logfile#统计61616未消费队列#grep -E --color 'size="[1-9][0-9]*"' activemq61616-data.xmlActiveMQ61616_Pending_Num=`grep -E --color 'size="[1-9][0-9]*"' $activemq61616PathXml |wc -l`echo "统计61616未消费队列[ActiveMQ61616_Pending_Num:$ActiveMQ61616_Pending_Num]" | tee -a $logfile#统计61616死信队列ActiveMQ61616_DLQ_Num=`grep -C 2 -E --color '<queue name="ActiveMQ.DLQ">' $activemq61616PathXml |grep  -E --color 'size="[1-9][0-9]*"' |wc -l`echo "统计61616死信队列[ActiveMQ61616_DLQ_Num:$ActiveMQ61616_DLQ_Num]" | tee -a $logfileif [ $ActiveMQ61616_Pending_Num = 0 ] || [ $ActiveMQ61616_Pending_Num = $ActiveMQ61616_DLQ_Num ]; then#没有未消费队列或者只有死信队列，执行61616归档check61616PendingMessagesResult=1;else#有未消费队列，无法执行61616归档，输出告警check61616PendingMessagesResult=2;fifiecho "完成确认61616是否有待消费(Number Of Pending Messages)MQ消息![data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}][check61616PendingMessagesResult:$check61616PendingMessagesResult]" | tee -a $logfile#未消费消息检查结果0-下载队列xml文件错误，1-没有未消费队列或者只有死信队列，可执行归档，2-有未消费队列，不执行归档check61626PendingMessagesResult=0;activemq61626PathXml="${logPath}/activemq61626-${cur_curl_datetime}.xml";activemq61626PathUrl="http://${check61626ServerIP}:8162/admin/xml/queues.jsp"curl -u $data_keep_xml_user:$data_keep_xml_password -o $activemq61626PathXml $activemq61626PathUrl#确认下载61626队列xml文件格式正常curl61626Result=`grep -E '<queues>|<queue name=' $activemq61626PathXml |wc -l`echo "curl61626Result:$curl61626Result" | tee -a $logfileif [ $curl61626Result = 0 ]; thenecho "下载61626队列xml文件错误${$activemq61626PathUrl}，无法确认未消费队列[curl61626Result:$curl61626Result]，此次定时归档失败！" | tee -a $logfileecho "start-----------------------------------------------------"  | tee -a $errorLogFileecho "　　　　　日志编号：监控脚本-2001" | tee -a $errorLogFileecho "　　　　　　　时间：$currTime" | tee -a $errorLogFileecho "　　　　　主机名称：$local_host($local_ip)" | tee -a $errorLogFileecho "　　　　　　　步骤：0" | tee -a $errorLogFileecho "　　　　　　　　类：ActiveMQ" | tee -a $errorLogFileecho "　　　　　功能描述：定时归档[周${data_keep_week},时${data_keep_hour}]" | tee -a $errorLogFileecho "　　　　　　用户名：SYS" | tee -a $errorLogFileecho "　　　　　　　日志：下载61626队列xml文件错误${activemq61626PathUrl}，无法确认未消费队列[curl61626Result:$curl61626Result]，此次定时归档失败！" | tee -a $errorLogFileecho "　　　　　解决方案：确认ActiveMQ-8162web管理服务正常。" | tee -a $errorLogFileecho "　　　　　　　级别：ERROR" | tee -a $errorLogFileecho "　　　　　其他参数：activemq61626PathXml:${activemq61626PathXml},data_keep_xml_user:${data_keep_xml_user},data_keep_xml_password:${data_keep_xml_password}]" | tee -a $errorLogFileecho "end----------------------------------------------------" | tee -a $errorLogFileelseecho "下载61626队列xml文件正常${activemq61626PathUrl}，[curl61626Result:$curl61626Result]" | tee -a $logfile#统计61626未消费队列#grep -E --color 'size="[1-9][0-9]*"' activemq61626-data.xmlActiveMQ61626_Pending_Num=`grep -E --color 'size="[1-9][0-9]*"' $activemq61626PathXml |wc -l`echo "统计61626未消费队列[ActiveMQ61626_Pending_Num:$ActiveMQ61626_Pending_Num]" | tee -a $logfile#统计61626死信队列ActiveMQ61626_DLQ_Num=`grep -C 2 -E --color '<queue name="ActiveMQ.DLQ">' $activemq61626PathXml |grep  -E --color 'size="[1-9][0-9]*"' |wc -l`echo "统计61626死信队列[ActiveMQ61626_DLQ_Num:$ActiveMQ61626_DLQ_Num]" | tee -a $logfileif [ $ActiveMQ61626_Pending_Num = 0 ] || [ $ActiveMQ61626_Pending_Num = $ActiveMQ61626_DLQ_Num ]; then#没有未消费队列或者只有死信队列，执行61626归档check61626PendingMessagesResult=1;else#有未消费队列，无法执行61626归档，输出告警check61626PendingMessagesResult=2;fifiecho "完成确认61626是否有待消费(Number Of Pending Messages)MQ消息![data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}][check61626PendingMessagesResult:$check61626PendingMessagesResult]" | tee -a $logfile#确认是否有待消费MQ消息end#记录本次执行归档结果data_keep_61616_state_now=$data_keep_61616_state;data_keep_61626_state_now=$data_keep_61626_state;for itemServer in $remote_hostsdoping -c2 -i0.3 -W1 $itemServer &>/dev/nullif [ $? -ne 0 ]; thenecho "${itemServer}，无法ping通!" | tee -a $logfileelseecho "开始执行归档和清理begin" | tee -a $logfileif [ $data_keep_61616_state -ne 1 ]; thenecho "开始结束进程：${itemServer}:${P1}" | tee -a $logfilelastDateTime=`date -d "${data_keep_61616_last_date}" +%s`;nowDate=$(date +"%Y-%m-%d %T");nowDateTime=`date -d "${nowDate}" +%s`;CountSeconds=$(($nowDateTime-$lastDateTime));CountDays=0;if [ $CountSeconds -gt 86400 ]; thenCountDays=`expr $CountSeconds / 86400`;fiecho "61616距离上次归档天数：[CountSeconds:${CountSeconds}，CountDays:${CountDays}]" | tee -a $logfile#可执行归档或者离上次成功归档超过1个月(30天)if [ $check61616PendingMessagesResult = 1 ] || [ $CountDays -ge $data_keep_day_count ]; thenstopresult=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "ps -ef |grep "$P1" | grep -v grep| awk '{print \$2}'| xargs kill -9 >/dev/null 2>&1 &")data_keep_61616_last_date=$(date +"%Y-%m-%d %T")data_keep_61616_state_now=1;echo "${data_keep_61616_last_date}完成结束进程！stopresult:${stopresult}" | tee -a $logfileelsedata_keep_61616_state_now=0;echo "未执行结束进程，8161有待消费MQ消息或者下载数据异常：${itemServer}:${P1}" | tee -a $logfilefielseecho "61616今天归档成功，不需要重复归档！[data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfilefiif [ $data_keep_61626_state -ne 1 ]; thenecho "开始结束进程：${itemServer}:${P2}" | tee -a $logfilelastDateTime=`date -d "${data_keep_61626_last_date}" +%s`;nowDate=$(date +"%Y-%m-%d %T");nowDateTime=`date -d "${nowDate}" +%s`;CountSeconds=$(($nowDateTime-$lastDateTime));CountDays=0;if [ $CountSeconds -gt 86400 ]; thenCountDays=`expr $CountSeconds / 86400`;fiecho "61626距离上次归档天数：[CountSeconds:${CountSeconds}，CountDays:${CountDays}]" | tee -a $logfile#可执行归档或者离上次成功归档超过1个月(30天)if [ $check61626PendingMessagesResult = 1 ] || [ $CountDays -ge $data_keep_day_count ]; thenstopresult2=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "ps -ef |grep "$P2" | grep -v grep| awk '{print \$2}'| xargs kill -9 >/dev/null 2>&1 &")data_keep_61626_last_date=$(date +"%Y-%m-%d %T");data_keep_61626_state_now=1;echo "${data_keep_61626_last_date}完成结束进程！stopresult2:${stopresult2}" | tee -a $logfileelsedata_keep_61626_state_now=0;echo "未执行结束进程，8162有待消费MQ消息或者下载数据异常：${itemServer}:${P2}" | tee -a $logfilefielseecho "61626今天归档成功，不需要重复归档！[data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfilefiif [ $data_keep_61616_state -ne 1 ]; thenecho "开始清理：${itemServer}:${P1}" | tee -a $logfile#find /mpjava/activemq-cluster1/ -mtime +7 -type d -name "data.bak*" -exec rm -rf {} \;rmresult=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "find "$P1" -mtime +7 -type d -name 'data.bak*' -exec rm -rf {} \;")echo "rmresult:${rmresult}" | tee -a $logfileelseecho "61616今天清理成功，不需要再清理！[data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfilefiif [ $data_keep_61626_state -ne 1 ]; thenecho "开始清理：${itemServer}:${P2}" | tee -a $logfile#find /mpjava/activemq-cluster2/ -mtime +7 -type d -name "data.bak*" -exec rm -rf {} \;rmresult2=$(sshpass -p "${remote_host_right_password}" ssh -o StrictHostKeyChecking=no $remote_host_user@$itemServer "find "$P2" -mtime +7 -type d -name 'data.bak*' -exec rm -rf {} \;")echo "rmresult2:${rmresult2}" | tee -a $logfileelseecho "61626今天清理成功，不需要再清理！[data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfilefiecho "完成执行归档和清理end" | tee -a $logfilefi done#更新本次执行归档结果data_keep_61616_state=$data_keep_61616_state_now;data_keep_61626_state=$data_keep_61626_state_now;#清理7天前/mpjava/amqportwath/*.log和*.xml文件echo "清理7天前${logPath}*.log和*.xml文件" | tee -a $logfilefind ${logPath} -mtime +7 -type f -name "*.log" -exec rm -rf {} \;find ${logPath} -mtime +7 -type f -name "*.xml" -exec rm -rf {} \;#如果61616或者61626归档失败，10分钟后重试if [ $data_keep_61616_state -ne 1 ] || [ $data_keep_61626_state -ne 1 ]; thendata_keep_state=0;#if [ $data_keep_state = 1 ]; then if [ $data_keep_61616_state -ne 1 ]; then #61616归档失败echo "61616归档失败！" | tee -a $logfileecho "start-----------------------------------------------------"  | tee -a $errorLogFileecho "　　　　　日志编号：监控脚本-2001" | tee -a $errorLogFileecho "　　　　　　　时间：$currTime" | tee -a $errorLogFileecho "　　　　　主机名称：$local_host($local_ip)" | tee -a $errorLogFileecho "　　　　　　　步骤：0" | tee -a $errorLogFileecho "　　　　　　　　类：ActiveMQ" | tee -a $errorLogFileecho "　　　　　功能描述：定时归档[周${data_keep_week},时${data_keep_hour}]" | tee -a $errorLogFileecho "　　　　　　用户名：SYS" | tee -a $errorLogFileecho "　　　　　　　日志：61616归档失败！原因是有待消费MQ消息或者下载数据异常。" | tee -a $errorLogFileecho "　　　　　解决方案：非生产时间，手动执行归档(结束三台MQ服务器的6个activemq进程)。" | tee -a $errorLogFileecho "　　　　　　　级别：ERROR" | tee -a $errorLogFileecho "　　　　　其他参数：" | tee -a $errorLogFileecho "end----------------------------------------------------" | tee -a $errorLogFilefiif [ $data_keep_61626_state -ne 1 ]; then #61626归档失败echo "61626归档失败！" | tee -a $logfileecho "start-----------------------------------------------------"  | tee -a $errorLogFileecho "　　　　　日志编号：监控脚本-2001" | tee -a $errorLogFileecho "　　　　　　　时间：$currTime" | tee -a $errorLogFileecho "　　　　　主机名称：$local_host($local_ip)" | tee -a $errorLogFileecho "　　　　　　　步骤：0" | tee -a $errorLogFileecho "　　　　　　　　类：ActiveMQ" | tee -a $errorLogFileecho "　　　　　功能描述：定时归档[周${data_keep_week},时${data_keep_hour}]" | tee -a $errorLogFileecho "　　　　　　用户名：SYS" | tee -a $errorLogFileecho "　　　　　　　日志：61626归档失败！原因是有待消费MQ消息或者下载数据异常。" | tee -a $errorLogFileecho "　　　　　解决方案：非生产时间，手动执行归档(结束三台MQ服务器的6个activemq进程)。" | tee -a $errorLogFileecho "　　　　　　　级别：ERROR" | tee -a $errorLogFileecho "　　　　　其他参数：" | tee -a $errorLogFileecho "end----------------------------------------------------" | tee -a $errorLogFilefi#fiecho "61616或者61626归档失败，10分钟后重试。[data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfileecho "sleep:10分钟(600s)" | tee -a $logfilesleep 600fielseecho "${currTime}今天已完成归档![周${data_keep_week},时${data_keep_hour}][data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfileecho "sleep:10分钟(600s)" | tee -a $logfilesleep 600fielseecho "${currTime}非归档时间![周${data_keep_week},时${data_keep_hour}][data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfile#归档状态复原data_keep_61616_state=0;data_keep_61626_state=0;data_keep_state=0;echo "${currTime}归档状态复原![data_keep_state：${data_keep_state}，data_keep_61616_state：${data_keep_61616_state}，data_keep_61626_state：${data_keep_61626_state}]" | tee -a $logfilefifi#AMQ定期归档end if [ $data_keep_state = 0 ]; then #AMQ端口61616/61626检查begincheck61616PortReslut=0check61616ServerIP=""checkAMQPort 61616 $P1check61616PortReslut=$checkPortReslutcheck61616ServerIP=$checkPortServerIPecho "check61616PortReslut${check61616PortReslut}" | tee -a $logfileecho "check61616ServerIP${check61616ServerIP}" | tee -a $logfilecheck61626PortReslut=0check61626ServerIP=""checkAMQPort 61626 $P2check61626PortReslut=$checkPortReslutcheck61626ServerIP=$checkPortServerIPecho "check61626PortReslut：${check61626PortReslut}" | tee -a $logfileecho "check61626ServerIP${check61626ServerIP}" | tee -a $logfile#AMQ端口61616/61626检查end if [ $check61616PortReslut -ne 0 ] && [ $check61626PortReslut -ne 0 ]; then#检查端口正常，每次循环沉睡1分钟(60s)echo "sleep:1分钟(60s)" | tee -a $logfilesleep 60else#检查端口不正常，结束进程后，20分钟再进行确认echo "sleep:20分钟(1200s)" | tee -a $logfilesleep 1200fifi
done

---

注意事项

1. 防火墙配置
   防火墙需开放以下端口：

   • 4369 (epmd)
   • 5672-5673 (AMQP)
   • 15672 (HTTP API)
   • 25672 (Erlang分发)

2. TLS加密
   建议在生产环境中使用TLS加密通信。可以使用以下命令生成证书：

   openssl req -x509 -newkey rsa:4096 -keyout key.pem -out cert.pem -days 365 -nodes
   

3. 定期清理策略
   设置队列的自动清理策略：

   rabbitmqctl set_policy auto-expire "^log.*" '{"expires":1800000}' --apply-to queues
   

---

完整部署包及配置文件获取

可以通过以下链接下载RabbitMQ的完整部署包及配置文件：

wget https://github.com/rabbitmq/rabbitmq-server/releases/download/v3.12.2/rabbitmq-server-generic-unix-3.12.2.tar.xz

---

通过以上步骤，您可以快速搭建一个高可用的RabbitMQ集群。希望本文对您有所帮助！如果有任何问题，欢迎在评论区留言。