通过Spring集成和RabbitMQ实现的高可用性AMQP-问题

我所做的：

[
 {rabbit, [ {tcp_listeners, [5673]},
 {collect_statistics_interval, 10000},
 {heartbeat,30},
 {cluster_partition_handling, pause_minority},
 {cluster_nodes, {[ 'rabbit@master',
                    'rabbit2@master'],
                  disc}} ] },
 {rabbitmq_management, [ {http_log_dir,"/tmp/rabbit-mgmt"},{listener, [{port, 15673}]} ] },
 {rabbitmq_management_agent, [ {force_fine_statistics, true} ] }
].

#!/bin/bash
echo "*** Second RabbitMQ Server is setting up ***"

export RABBITMQ_HOSTNAME=rabbit2@master
export RABBITMQ_NODE_PORT=5673
export RABBITMQ_NODENAME=rabbit2@master
export RABBITMQ_SERVER_START_ARGS="-rabbitmq_management listener [{port,15673}]"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmq-server &

echo "*** Second RabbitMQ Server is set up succesfully. ***"

sleep 5

echo "*** Second RabbitMQ Server' s status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl status

sleep 5

echo "*** Second RabbitMQ Server is being added to cluster... ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master stop_app
/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master join_cluster rabbit@master
/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master start_app
/DEV_TOOLS/rabbitmq_server-3.4.2/sbin/rabbitmqctl -n rabbit@master set_policy ha-all "^ha\." '{"ha-mode":"all"}'

echo "*** Second RabbitMQ Server is added to cluster successfully... ***"

sleep 5

echo "*** Second RabbitMQ Server' s cluster status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl cluster_status

我正在使用RabbitMQ消息代理，它在我的spring micro服务中支持AMQP标准。其主要目的是在rabbitMQ服务器（不幸）关闭期间防止消息丢失。因此，我计划提供高度可用的AMQP支持的消息通道

我尝试设置两个RabbitMQ服务器，并按照下面的链接设置RabbitMQ服务器

第一个RabbitMQ服务器的配置文件（RabbitMQ.config）如下所示。它应该放在../rabbitmq_server-version/etc/rabbitmq下/

[
 {rabbit, [ {tcp_listeners, [5672]},
 {collect_statistics_interval, 10000},
 {heartbeat,30},
 {cluster_partition_handling, pause_minority},
 {cluster_nodes, {[ 'rabbit@master',
                    'rabbit2@master'],
                  disc}} ] },
 {rabbitmq_management, [ {http_log_dir,"/tmp/rabbit-mgmt"},{listener, [{port, 15672}]} ] },
 {rabbitmq_management_agent, [ {force_fine_statistics, true} ] }
].

第二个RabbitMQ服务器的RabbitMQ.config文件：

[
 {rabbit, [ {tcp_listeners, [5673]},
 {collect_statistics_interval, 10000},
 {heartbeat,30},
 {cluster_partition_handling, pause_minority},
 {cluster_nodes, {[ 'rabbit@master',
                    'rabbit2@master'],
                  disc}} ] },
 {rabbitmq_management, [ {http_log_dir,"/tmp/rabbit-mgmt"},{listener, [{port, 15673}]} ] },
 {rabbitmq_management_agent, [ {force_fine_statistics, true} ] }
].

#!/bin/bash
echo "*** Second RabbitMQ Server is setting up ***"

export RABBITMQ_HOSTNAME=rabbit2@master
export RABBITMQ_NODE_PORT=5673
export RABBITMQ_NODENAME=rabbit2@master
export RABBITMQ_SERVER_START_ARGS="-rabbitmq_management listener [{port,15673}]"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmq-server &

echo "*** Second RabbitMQ Server is set up succesfully. ***"

sleep 5

echo "*** Second RabbitMQ Server' s status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl status

sleep 5

echo "*** Second RabbitMQ Server is being added to cluster... ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master stop_app
/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master join_cluster rabbit@master
/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master start_app
/DEV_TOOLS/rabbitmq_server-3.4.2/sbin/rabbitmqctl -n rabbit@master set_policy ha-all "^ha\." '{"ha-mode":"all"}'

echo "*** Second RabbitMQ Server is added to cluster successfully... ***"

sleep 5

echo "*** Second RabbitMQ Server' s cluster status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl cluster_status

第一个RabbitMQ服务器的bash脚本示例如下所示。另外，请查看RabbitMQ集群文档以了解其他配置步骤

#!/bin/bash
echo "*** First RabbitMQ Server is setting up ***"

export RABBITMQ_HOSTNAME=rabbit@master
export RABBITMQ_NODE_PORT=5672
export RABBITMQ_NODENAME=rabbit@master
export RABBITMQ_SERVER_START_ARGS="-rabbitmq_management listener [{port,15672}]"

/DEV_TOOLS/rabbitmq_server-3.4.2/sbin/rabbitmq-server &

echo "*** Second RabbitMQ Server is set up succesfully. ***"

sleep 5

echo "*** First RabbitMQ Server' s status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2/sbin/rabbitmqctl status

第二个RabbitMQ服务器的bash脚本示例如下：

[
 {rabbit, [ {tcp_listeners, [5673]},
 {collect_statistics_interval, 10000},
 {heartbeat,30},
 {cluster_partition_handling, pause_minority},
 {cluster_nodes, {[ 'rabbit@master',
                    'rabbit2@master'],
                  disc}} ] },
 {rabbitmq_management, [ {http_log_dir,"/tmp/rabbit-mgmt"},{listener, [{port, 15673}]} ] },
 {rabbitmq_management_agent, [ {force_fine_statistics, true} ] }
].

#!/bin/bash
echo "*** Second RabbitMQ Server is setting up ***"

export RABBITMQ_HOSTNAME=rabbit2@master
export RABBITMQ_NODE_PORT=5673
export RABBITMQ_NODENAME=rabbit2@master
export RABBITMQ_SERVER_START_ARGS="-rabbitmq_management listener [{port,15673}]"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmq-server &

echo "*** Second RabbitMQ Server is set up succesfully. ***"

sleep 5

echo "*** Second RabbitMQ Server' s status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl status

sleep 5

echo "*** Second RabbitMQ Server is being added to cluster... ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master stop_app
/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master join_cluster rabbit@master
/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master start_app
/DEV_TOOLS/rabbitmq_server-3.4.2/sbin/rabbitmqctl -n rabbit@master set_policy ha-all "^ha\." '{"ha-mode":"all"}'

echo "*** Second RabbitMQ Server is added to cluster successfully... ***"

sleep 5

echo "*** Second RabbitMQ Server' s cluster status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl cluster_status

问题是什么：

[
 {rabbit, [ {tcp_listeners, [5673]},
 {collect_statistics_interval, 10000},
 {heartbeat,30},
 {cluster_partition_handling, pause_minority},
 {cluster_nodes, {[ 'rabbit@master',
                    'rabbit2@master'],
                  disc}} ] },
 {rabbitmq_management, [ {http_log_dir,"/tmp/rabbit-mgmt"},{listener, [{port, 15673}]} ] },
 {rabbitmq_management_agent, [ {force_fine_statistics, true} ] }
].

#!/bin/bash
echo "*** Second RabbitMQ Server is setting up ***"

export RABBITMQ_HOSTNAME=rabbit2@master
export RABBITMQ_NODE_PORT=5673
export RABBITMQ_NODENAME=rabbit2@master
export RABBITMQ_SERVER_START_ARGS="-rabbitmq_management listener [{port,15673}]"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmq-server &

echo "*** Second RabbitMQ Server is set up succesfully. ***"

sleep 5

echo "*** Second RabbitMQ Server' s status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl status

sleep 5

echo "*** Second RabbitMQ Server is being added to cluster... ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master stop_app
/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master join_cluster rabbit@master
/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl -n rabbit2@master start_app
/DEV_TOOLS/rabbitmq_server-3.4.2/sbin/rabbitmqctl -n rabbit@master set_policy ha-all "^ha\." '{"ha-mode":"all"}'

echo "*** Second RabbitMQ Server is added to cluster successfully... ***"

sleep 5

echo "*** Second RabbitMQ Server' s cluster status : ***"

/DEV_TOOLS/rabbitmq_server-3.4.2_2/sbin/rabbitmqctl cluster_status

我可以看到两个RabbitMQ服务器正在同步运行

当我停止第一个节点时，第二个节点正在工作

sbin/rabbitmqctl stop_app

但是当我关闭第一个RabbitMQ服务器时，第二个节点也会关闭

sbin/rabbitmqctl stop

但是结果应该是,

如果第一个消息传递节点和第一个RabbitMQ服务器关闭 意外地，第二个消息传递节点和第二个RabbitMQ服务器将 继续处理订单消息，以防止潜在的消息丢失和 使用高可用性可以防止服务中断问题 AMQP支持的通道

---

请帮我解决这个问题。谢谢。

这个问题实际上与您引用的spring项目无关。如果将纯rabbitmq问题提交给rabbitmq用户google组，你可能会更走运——这就是rabbitmq工程师们常去的地方。这个问题实际上与你引用的spring项目毫无关系。通过将纯rabbitmq问题引导到rabbitmq用户google组，您可能会有更好的运气——这就是rabbitmq工程师经常出现的地方。