├── dockerfiles ├── erlang.cookie ├── Dockerfile ├── startrabbit.sh └── rabbitmq.config ├── imgs ├── 1.png ├── rpc.png ├── seq.png ├── amqp.png ├── policies.png ├── routing.png ├── topics.png ├── broadcast.png ├── management.png ├── workqueue.png ├── performance.png ├── federated_cluster.png ├── message-chains-1.png └── federated_broadcast.png ├── docker_rabbitmq.sh ├── README.md └── rabbitmq使用心得.md /dockerfiles/erlang.cookie: -------------------------------------------------------------------------------- 1 | ERLANGCOOKIE -------------------------------------------------------------------------------- /imgs/1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/1.png -------------------------------------------------------------------------------- /imgs/rpc.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/rpc.png -------------------------------------------------------------------------------- /imgs/seq.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/seq.png -------------------------------------------------------------------------------- /imgs/amqp.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/amqp.png -------------------------------------------------------------------------------- /imgs/policies.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/policies.png -------------------------------------------------------------------------------- /imgs/routing.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/routing.png -------------------------------------------------------------------------------- /imgs/topics.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/topics.png -------------------------------------------------------------------------------- /imgs/broadcast.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/broadcast.png -------------------------------------------------------------------------------- /imgs/management.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/management.png -------------------------------------------------------------------------------- /imgs/workqueue.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/workqueue.png -------------------------------------------------------------------------------- /imgs/performance.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/performance.png -------------------------------------------------------------------------------- /imgs/federated_cluster.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/federated_cluster.png -------------------------------------------------------------------------------- /imgs/message-chains-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/message-chains-1.png -------------------------------------------------------------------------------- /imgs/federated_broadcast.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/HEAD/imgs/federated_broadcast.png -------------------------------------------------------------------------------- /dockerfiles/Dockerfile: -------------------------------------------------------------------------------- 1 | FROM rabbitmq:3.5.3 2 | MAINTAINER slv slv@ilegendsoft.com 3 | 4 | RUN mkdir /opt/rabbit 5 | ADD startrabbit.sh /opt/rabbit/ 6 | ADD rabbitmq.config /etc/rabbitmq/ 7 | ADD erlang.cookie /var/lib/rabbitmq/.erlang.cookie 8 | 9 | RUN chmod a+x /opt/rabbit/startrabbit.sh 10 | RUN chmod u+rw /etc/rabbitmq/rabbitmq.config 11 | 12 | RUN rabbitmq-plugins enable --offline rabbitmq_mqtt rabbitmq_stomp rabbitmq_management rabbitmq_management_agent rabbitmq_management_visualiser rabbitmq_federation rabbitmq_federation_management sockjs 13 | 14 | EXPOSE 5672 15 | EXPOSE 15672 16 | EXPOSE 25672 17 | EXPOSE 4369 18 | EXPOSE 9100 19 | EXPOSE 9101 20 | EXPOSE 9102 21 | EXPOSE 9103 22 | EXPOSE 9104 23 | EXPOSE 9105 24 | 25 | CMD /opt/rabbit/startrabbit.sh -------------------------------------------------------------------------------- /docker_rabbitmq.sh: -------------------------------------------------------------------------------- 1 | #!/bin/sh 2 | IMAGE=ilegendsoft/rabbitmq-server-cluster 3 | 4 | docker pull $IMAGE 5 | 6 | # if first server: 7 | docker run -d -p 5672:5672 -p 15672:15672 -e RABBITMQ_NODENAME=rabbit -e RABBITMQ_SERVER_START_ARGS="-rabbit cluster_nodes {['rabbit@host1','rabbit@host2'],disc}" --name rabbitmq-server-cluster --net="host" $IMAGE 8 | 9 | # elif cluster server on other host: 10 | #docker run -d -p 5672:5672 -p 15672:15672 -e RABBITMQ_NODENAME=rabbit -e CLUSTERED=true -e CLUSTER_WITH=zclouddev1 -e CLUSTER_NODENAME=rabbit --name rabbitmq-server-cluster --net="host" $IMAGE 11 | 12 | # else cluster server on same host: 13 | #docker run -d -p 5673:5673 -p 15673:15673 -e RABBITMQ_NODENAME=rabbit2 -e CLUSTERED=true -e CLUSTER_WITH=boot2docker -e CLUSTER_NODENAME=rabbit -e RABBITMQ_DIST_PORT=25673 -e RABBITMQ_NODE_PORT=5673 -e RABBITMQ_SERVER_START_ARGS="-rabbitmq_management listener [{port,15673}]" --name rabbitmq-server-cluster2 --net="host" $IMAGE -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # rabbitmq-server-cluster 2 | rabbitmq server cluster docker image 3 | 4 | ## introduction 5 | This is a project about how to build a rabbitmq cluster docker image. 6 | 7 | ## how to use it 8 | * ensure you have installed docker 9 | * `docker build -t ilegendsoft/rabbitmq-server-cluster dockerfiles/` 10 | * execute `docker_rabbitmq.sh` 11 | * `docker ps` or login [http://host1:15672/](http://host1:15672/) with admin/admin to check status. 12 | 13 | ## configuration 14 | * `dockerfiles/Dockerfile`:Dockerfile 15 | * `dockerfiles/erlang.cookie`:Rabbitmq cluster nodes' cookie need to be same. 16 | * `dockerfiles/rabbitmq.config`:some rabbitmq's configuration, such as user, memory, disk, and so on. 17 | * `dockerfiles/startrabbit.sh`:CMD in Dockerfile 18 | 19 | * `docker_rabbitmq.sh`:start the container 20 | `docker run -d -p 5672:5672 -p 15672:15672 -e RABBITMQ_NODENAME=rabbit -e RABBITMQ_USE_LONGNAME=true -e RABBITMQ_SERVER_START_ARGS="-rabbit cluster_nodes {['rabbit@host1','rabbit@host2'],disc}" --name rabbitmq-server-cluster --net="host" $IMAGE` 21 | 22 | 23 | ***host1,host2 are your cluster nodes' hostnames, you can add more than two.*** 24 | -------------------------------------------------------------------------------- /dockerfiles/startrabbit.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | chown rabbitmq:rabbitmq /var/lib/rabbitmq/.erlang.cookie 3 | chmod 400 /var/lib/rabbitmq/.erlang.cookie 4 | 5 | RABBITMQ_NODENAME=$RABBITMQ_NODENAME /usr/sbin/rabbitmq-server 6 | 7 | # if [ -z "$CLUSTERED" ]; then 8 | # # if not clustered then start it normally as if it is a single server 9 | # RABBITMQ_NODENAME=$RABBITMQ_NODENAME /usr/sbin/rabbitmq-server 10 | # else 11 | # if [ -z "$CLUSTER_WITH" -o -z "$CLUSTER_NODENAME" ]; then 12 | # # If clustered, but cluster with is not specified then again start normally, could be the first server in the 13 | # # cluster 14 | # RABBITMQ_NODENAME=$RABBITMQ_NODENAME /usr/sbin/rabbitmq-server 15 | # else 16 | # RABBITMQ_NODENAME=$RABBITMQ_NODENAME /usr/sbin/rabbitmq-server -detached 17 | # /usr/sbin/rabbitmqctl stop_app 18 | # if [ -z "$RAM_NODE" ]; then 19 | # /usr/sbin/rabbitmqctl join_cluster $CLUSTER_NODENAME@$CLUSTER_WITH 20 | # else 21 | # /usr/sbin/rabbitmqctl join_cluster --ram $CLUSTER_NODENAME@$CLUSTER_WITH 22 | # fi 23 | # /usr/sbin/rabbitmqctl start_app 24 | # # /usr/sbin/rabbitmq-plugins enable rabbitmq_mqtt rabbitmq_stomp rabbitmq_management rabbitmq_management_agent rabbitmq_management_visualiser rabbitmq_federation rabbitmq_federation_management sockjs 25 | 26 | # # Tail to keep the a foreground process active.. 27 | # #tail -f /var/log/rabbitmq/rabbit\@$HOSTNAME.log 28 | # tail -f /etc/hosts 29 | # fi 30 | # fi 31 | 32 | -------------------------------------------------------------------------------- /rabbitmq使用心得.md: -------------------------------------------------------------------------------- 1 | ![cover](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/message-chains-1.png) 2 | 3 | # RabbitMQ在分布式系统中的应用 4 | 5 | 由于之前做的项目中需要在多个节点之间**可靠地**通信，所以废弃了之前使用的Redis pub/sub（因为集群有单点问题，且有诸多限制），改用了RabbitMQ。 6 | 使用期间得到不少收获，也踩了不少坑，所以在此分享下心得。（简单了解下RabbitMQ? [点这里](#other)） 7 | 8 | ### 怎么保证可靠性的？ 9 | 10 | RabbitMQ提供了几种特性，牺牲了一点性能代价，提供了可靠性的保证。 11 | 12 | * 持久化 13 | 当RabbitMQ退出时，默认会将消息和队列都清除，所以需要在**第一次**声明队列和发送消息时指定其持久化属性为true，这样RabbitMQ会将队列、消息和状态存到RabbitMQ本地的数据库，重启后会恢复。 14 | java: 15 | 16 | durable=true 17 | channel.queueDeclare("task_queue", durable, false, false, null); // 队列 18 | channel.basicPublish("", "task_queue", 19 | MessageProperties.PERSISTENT_TEXT_PLAIN, 20 | message.getBytes()); // 消息 21 | 22 | 注：当声明的队列已经存在时，尝试重新定义它的durable是不生效的。 23 | 24 | * 接收应答 25 | 客户端接收消息的模式默认是自动应答，但是通过设置autoAck为false可以让客户端主动应答消息。当客户端拒绝此消息或者未应答便断开连接时，就会使得此消息重新入队（在版本2.7.0以前是到重新加入到队尾，2.7.0及以后是保留消息在队列中的原来位置）。 26 | java: 27 | 28 | autoAck = false; 29 | requeue = true; 30 | channel.basicConsume(queue, autoAck, callback); 31 | channel.basicAck();//应答 32 | channel.basicReject(deliveryTag, requeue); // 拒绝 33 | channel.basicRecover(requeue); // 恢复 34 | 35 | * 发送确认 36 | 默认情况下，发送端不关注发出去的消息是否被消费掉了。可设置channel为confirm模式，所有发送的消息都会被确认一次，用户可以自行根据server发回的确认消息查看状态。详细介绍见：[confirms](http://www.rabbitmq.com/confirms.html) 37 | java: 38 | 39 | channel.confirmSelect(); // 进入confirm模式 40 | // do publish messages... 每条消息都会被编号，从1开始 41 | channel.getNextPublishSeqNo() // 查看下一条要发送的消息的序号 42 | channel.waitForConfirms(); // 等待所有消息发送并确认 43 | 44 | * 事务：和confirm模式不能同时使用，而且会带来大量的多余开销，导致吞吐量下降很多，故而不推荐。 45 | java: 46 | 47 | channel.txSelect(); 48 | try { 49 | // do something... 50 | channel.txCommit(); 51 | } catch (e){ 52 | channel.txRollback(); 53 | } 54 | 55 | * 消息队列的高可用（主备模式） 56 | 相比于路由和绑定，可以视为是共享于所有的节点的，消息队列默认只存在于第一次声明它的节点上，这样一旦这个节点挂了，这个队列中未处理的消息就没有了。 57 |
幸好，RabbitMQ提供了将它备份到其他节点的机制，任何时候都有一个master负责处理请求，其他slaves负责备份，当master挂掉，会将最早创建的那个slave提升为master。 58 | 命令： 59 | `rabbitmqctl set_policy ha-all "^ha\." '{"ha-mode":"all"}'`：设置所有以'ha'开头的queue在所有节点上拥有备份。[详细语法点这里](http://www.rabbitmq.com/ha.html)； 60 | 也可以在界面上配置。 61 | 62 | 注：由于exclusive类型的队列会在client和server连接断开时被删掉，所以对它设置持久化属性和备份都是没有意义的。 63 | 64 | * 顺序保证 65 | 直接上图好了： 66 | ![seq](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/seq.png) 67 | 68 | ### 一些需要注意的地方 69 | 70 | * 集群配置： 71 | 一个集群中多个节点共享一份.erlang.cookie文件；若是没有启用RABBITMQ_USE_LONGNAME，需要在每个节点的hosts文件中指定其他节点的地址，不然会找不到其他集群中的节点。 72 | 73 | * 脑裂（网络分区）： 74 | RabbitMQ集群对于网络分区的处理和忍受能力不太好，推荐使用[federation](#federation)或者shovel插件去解决。 75 |
但是，情况已经发生了，怎么去解决呢？放心，还是有办法恢复的。 76 |
当网络断断续续时，会使得节点之间的通信断掉，进而造成集群被分隔开的情况。 77 |
这样，每个小集群之后便只处理各自本地的连接和消息，从而导致数据不同步。当重新恢复网络连接时，它们彼此都认为是对方挂了-_-||，便可以判断出有网络分区出现了。但是RabbitMQ默认是忽略掉不处理的，造成两个节点继续各自为政（路由，绑定关系，队列等可以独立地创建删除，甚至主备队列也会每一方拥有自己的master）。 78 |
可以更改配置使得连接恢复时，会根据配置自动恢复： 79 | 80 | * ignore：默认，不做任何处理 81 | * pause-minority：断开连接时，判断当前节点是否属于少数派（节点数少于或者等于一半），如果是，则暂停直到恢复连接。 82 | * {pause_if_all_down, [nodes], ignore | autoheal}：断开连接时，判断当前集群中节点是否有节点在nodes中，如果有，则继续运行，否则暂停直到恢复连接。这种策略下，当恢复连接时，可能会有多个分区存活，所以，最后一个参数决定它们怎么合并。 83 | * autoheal：当恢复连接时，选择客户端连接数最多的节点状态为主，重启其他节点。 84 | 85 | 配置：[集群配置](#cluster) 86 | 87 | * 多次ack 88 | 客户端多次应答同一条消息，会使得该客户端收不到后续消息。 89 | 90 | ### 结合Docker使用 91 | 集群版本的实现：详见我自己写的一个例子[rabbitmq-server-cluster](https://github.com/dengyuankai272/rabbitmq-server-cluster) 92 | 93 | ### 消息队列中间件的比较 94 | 95 | * RabbitMQ： 96 | 97 | * 优点：支持很多协议如：AMQP，XMPP，STMP，STOMP；灵活的路由；成熟稳定的集群方案；负载均衡；数据持久化等。 98 | * 缺点：速度较慢；比较重量级，安装需要依赖Erlang环境。 99 | 100 | * Redis： 101 | 102 | * 优点：比较轻量级，易上手 103 | * 缺点：单点问题，功能单一 104 | 105 | * Kafka： 106 | 107 | * 优点：高吞吐；分布式；快速持久化；负载均衡；轻量级 108 | * 缺点：极端情况下会丢消息 109 | 110 | 最后附一张网上截取的测试结果: 111 | 112 | ![performance](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/performance.png) 113 | 114 | 更多性能参数见：[http://www.rabbitmq.com/blog/2012/04/25/rabbitmq-performance-measurements-part-2/](http://www.rabbitmq.com/blog/2012/04/25/rabbitmq-performance-measurements-part-2/) 115 | 116 | 如果有兴趣简单了解下RabbitMQ的简单介绍，可以继续往下看～ 117 | 118 | # 简介 119 | 120 | ## 几个重要的概念 121 | 122 | * Virtual Host: 包含若干个Exchange和Queue，表示一个节点； 123 | * Exchange: 接受客户端发送的消息，并根据Binding将消息路由给服务器中的队列，Exchange分为direct, fanout, topic三种。 124 | * Binding: 连接Exchange和Queue，包含路由规则。 125 | * Queue: 消息队列，存储还未被消费的消息。 126 | * Message: Header+Body 127 | * Channel: 通道，执行AMQP的命令；一个连接可创建多个通道以节省资源。 128 | 129 | ## Client 130 | RabbitMQ官方实现了很多热门语言的客户端，就不一一列举啦，以java为例，直接开始正题： 131 | 132 | * 建立连接： 133 | 134 | ConnectionFactory factory = new ConnectionFactory(); 135 | factory.setHost("localhost"); 136 | 可以加上断开重试机制： 137 | 138 | factory.setAutomaticRecoveryEnabled(true); 139 | factory.setNetworkRecoveryInterval(10000); 140 | 创建连接和通道： 141 | 142 | Connection connection = factory.newConnection(); 143 | Channel channel = connection.createChannel(); 144 | 145 | * 一对一：一个生产者，一个消费者 146 | 147 | ![1](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/1.png) 148 | 149 | 生产者： 150 | 151 | channel.queueDeclare(QUEUE_NAME, false, false, false, null); 152 | channel.basicPublish("", QUEUE_NAME, null, message.getBytes()); 153 | 154 | 消费者： 155 | 156 | Consumer consumer = new DefaultConsumer(channel) { 157 | @Override 158 | public void handleDelivery(String consumerTag, Envelope envelope, AMQP.BasicProperties properties, byte[] body) 159 | throws IOException { 160 | String message = new String(body, "UTF-8"); 161 | System.out.println(" [x] Received '" + message + "'"); 162 | } 163 | }; 164 | channel.basicConsume(QUEUE_NAME, autoAck, consumer); 165 | 166 | * 一对多：一个生产者，多个消费者 167 | 168 | ![workqueue](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/workqueue.png) 169 | 170 | 代码同上，只不过会有多个消费者，消息会轮序发给各个消费者。 171 | 如果设置了autoAck=false，那么可以实现公平分发（即对于某个特定的消费者，每次最多只发送指定条数的消息，直到其中一条消息应答后，再发送下一条）。需要在消费者中加上: 172 | 173 | int prefetchCount = 1; 174 | channel.basicQos(prefetchCount); 175 | 其他同上。 176 | 177 | * 广播 178 | 179 | ![broadcast](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/broadcast.png) 180 | 181 | 生产者： 182 | 183 | channel.exchangeDeclare(EXCHANGE_NAME, "fanout"); 184 | String queueName = channel.queueDeclare().getQueue(); 185 | channel.queueBind(queueName, EXCHANGE_NAME, ""); 186 | channel.basicPublish(EXCHANGE_NAME, "", null, message.getBytes()); 187 | 188 | 消费者同上。 189 | 190 | * Routing: 指定路由规则 191 | 192 | ![routing](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/routing.png) 193 | 194 | 生产者： 195 | 196 | String queueName = channel.queueDeclare().getQueue(); 197 | channel.queueBind(queueName, EXCHANGE_NAME, routingKey); 198 | channel.basicPublish(EXCHANGE_NAME, routingKey, null, message.getBytes()); 199 | 消费者同上。 200 | 201 | * Topics: 支持通配符的Routing 202 | 203 | ![topics](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/topics.png) 204 | 205 | *可以表示一个单词 206 | #可以表示一个或多个单词 207 | 208 | 生产者： 209 | 210 | channel.exchangeDeclare(EXCHANGE_NAME, "topic"); 211 | String queueName = channel.queueDeclare().getQueue(); 212 | channel.queueBind(queueName, EXCHANGE_NAME, bindingKey); 213 | 消费者同上。 214 | 215 | * RPC 216 | 217 | ![rpc](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/rpc.png) 218 | 219 | 其实就是一对一模式的一种用法： 220 | 首先，客户端发送一条消息到服务端声明的队列，消息属性中包含reply_to和correlation_id 221 | 222 | - reply_to 是客户端创建的消息的队列，用来接收远程调用结果 223 | - correlation_id 是消息的标识，服务端回应的消息属性中会带上以便知道是哪条消息的结果。 224 | 225 | 然后，服务端接收到消息，处理，并返回一条结果到reply_to队列中， 226 | 227 | 最终，客户端接收到返回消息，继续向下处理。 228 | 229 | 230 | ## Server 231 | 支持各大主流操作系统，这里以Unix为例介绍下常用配置和命令： 232 | 233 | ### 安装 234 | 由于RabbitMQ是依赖于Erlang的，所以得首先安装最近版本的Erlang。 235 | 236 | 单点的安装比较简单，下载解压即可。[下载地址](http://www.rabbitmq.com/download.html) 237 | 238 | * 配置：（一般的，用默认的即可。） 239 | * $RABBITMQ_HOME/etc/rabbitmq/rabbitmq-env.conf: 环境变量默认配置（也可在启动脚本中设置，且以启动命令中的配置为准）。常用的有： 240 | * RABBITMQ_NODENAME：节点名称，默认是rabbit@$HOSTNAME。 241 | * RABBITMQ_NODE_PORT：协议端口号，默认5672。 242 | * RABBITMQ_SERVER_START_ARGS：覆盖rabbitmq.config中的一些配置。 243 | 244 | * $RABBITMQ_HOME/etc/rabbitmq/rabbitmq.config: 核心组件，插件，erlang服务等配置，常用的有： 245 | * disk_free_limit：队列持久化等信息都是存到RabbitMQ本地的数据库中的，默认限制50000000（也就是最多只让它使用50M空间啦，不够可以上调，也支持空闲空间百分比的配置）。要是超标了，它就罢工了…… 246 | * vm_memory_high_watermark：内存使用，默认0.4（最多让它使用40%的内存，超标罢工） 247 | 248 | 注：若启动失败了，可以在启动日志中查看到具体的错误信息。 249 | 250 | * 命令： 251 | * `$RABBITMQ_HOME/sbin/rabbitmq-server`：启动脚本，会打印出配置文件，插件，集群等信息；加上-detached为后台启动； 252 | * `/sbin/rabbitmqctl status`：查看启动状态 253 | * `/sbin/rabbitmqctl add_user admin admin`：添加新用户admin，密码admin；默认只有一个guest用户，但只限本机访问。 254 | * `/sbin/rabbitmqctl set_user_tags admin administrator`：将admin设置为管理员权限 255 | * `/sbin/rabbitmqctl set_permissions -p / admin ".*" ".*" ".*"` 赋予admin所有权限 256 | * `/sbin/rabbitmqctl stop`：关闭 257 | 258 | ### 集群 259 | 260 | 集群节点共享所有的状态和数据，如：用户、路由、绑定等信息（队列有点特殊，虽然从所有节点都可达，但是只存在于第一次声明它的那个节点上，解决方案：[消息队列的高可用](#ha)）；每个节点都可以接收连接，处理数据。 261 | 262 | 集群节点有两种，disc：默认，信息存在本地数据库；ram：加入集群时，添加--ram参数，信息存在内存，可提高性能。 263 | 264 | * 配置：（一般的，用默认的即可。） 265 | * $RABBITMQ_HOME/etc/rabbitmq/rabbitmq-env.conf： 266 | * RABBITMQ_USE_LONGNAME：默认false，(默认的，RABBITMQ_NODENAME中@后面的$HOSTNAME是主机名，所以需要集群中每个节点的hosts文件包含其他节点主机名到地址的映射。但是如果设置为true，就可以定义RABBITMQ_NODENAME中的$HOSTNAME为域名了） 267 | * RABBITMQ_DIST_PORT：集群端口号，默认RABBITMQ_NODE_PORT + 20000 268 | * $RABBITMQ_HOME/etc/rabbitmq/rabbitmq.config： 269 | * cluster_nodes：设置后，在启动时会尝试自动连接加入的节点并组成集群。 270 | * cluster_partition_handling：[网络分区的处理](#cluster_partion)。 271 | 272 | 更多详细的配置见：[配置](http://www.rabbitmq.com/configure.html) 273 | 274 | * 命令 275 | * `rabbitmqctl stop_app` 276 | * `rabbitmqctl join_cluster [--ram] nodename@hostname`：将当前节点加入到集群中；默认是以disc节点加入集群，加上--ram为ram节点。 277 | * `rabbitmqctl start_app` 278 | * `rabbitmqctl cluster_status`：查看集群状态 279 | 280 | 注：如果加入集群失败，可先查看 281 | 282 | * 每个节点的$HOME/.erlang.cookie内容一致； 283 | * 如果hostname是主机名，那么此hostname和地址的映射需要加入hosts文件中； 284 | * 如果使用的是域名，那么需要设置RABBITMQ_USE_LONGNAME为true。 285 | 286 | 注：docker版集群的见：[rabbitmq-server-cluster](https://github.com/dengyuankai272/rabbitmq-server-cluster) 287 | 288 | ## 高级 289 | 290 | ### AMQP协议简介 291 | RabbitMQ原生支持AMQP 0-9-1并扩展实现了了一些常用的功能：[AMQP 0-9-1](http://www.rabbitmq.com/amqp-0-9-1-reference.html) 292 | 293 | 包含三层： 294 | 295 | * 模型层: 最高层，提供了客户端调用的命令，如：queue.declare,basic.ack,consume等。 296 | * 会话层：将命令从客户端传递给服务器，再将服务器的应答传递给客户端，会话层为这个传递过程提供可靠性、同步机制和错误处理。 297 | * 传输层：主要传输二进制数据流，提供帧的处理、信道复用、错误检测和数据表示。 298 | 299 | ![](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/amqp.png) 300 | 301 | 注：其他协议的支持见：[RabbitMQ支持的协议](http://www.rabbitmq.com/protocols.html) 302 | 303 | ### 常用插件 304 | 305 | #### 管理界面（神器） 306 | 启动后，执行`rabbitmq-plugins enable rabbitmq_management`-> 307 | 访问[http://localhost:15672](http://localhost:15672)->查看节点状态，队列信息等等，甚至可以动态配置消息队列的主备策略，如下图： 308 | 309 | ![management](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/management.png) 310 | 311 | #### Federation 312 | 313 | 启用Federation插件，使得不同集群的节点之间可以传递消息，从而模拟出类似集群的效果。这样可以有几点好处： 314 | 315 | * 松耦合：联合在一起的不同集群可以有各自的用户，权限等信息，无需一致；此外，这些集群的RabbitMQ和Erlang的版本可以不一致。 316 | * 远程网络连接友好：由于通信是遵循AMQP协议的，故而对断断续续的网络连接容忍度高。 317 | * 自定义：可以自主选择哪些组件启用federation。 318 | 319 | 几个概念： 320 | 321 | * Upstreams: 定义上游节点信息，如： 322 | `rabbitmqctl set_parameter federation-upstream my-upstream '{"uri":"amqp://server-name","expires":3600000}'` 定义一个my-upstream 323 | * uri是其上游节点的地址，多个upstream的节点无需在同一集群中。 324 | * expires表示断开连接3600000ms后其上游节点会缓存消息。 325 | 326 | * Upstream sets: 多个Upstream的集合；默认有个`all`，会将所有的Upstream加进去。 327 | 328 | * Policies: 定义哪些exchanges,queues关联到哪个Upstream或者Upstream set，如： 329 | `rabbitmqctl set_policy --apply-to exchanges federate-me "^amq\." '{"federation-upstream-set":"all"}'` 将此节点所有以amq.开头的exchange联合到上游节点的同名exchange。 330 | 331 | 注： 332 | 333 | * 由于下游节点的exchange可以继续作为其他节点的上游，故可设置成循环，广播等形式。 334 | * 通过max_hops参数控制传递层数。 335 | * 模拟集群，可以将多个节点两两互连，并设置max_hops=1。 336 | 337 | ![federated_cluster](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/federated_cluster.png) 338 | ![federated_broadcast](https://raw.githubusercontent.com/dengyuankai272/rabbitmq-server-cluster/develop/imgs/federated_broadcast.png) 339 | 340 | `rabbitmq-plugins enable rabbitmq_federation` 341 | 如果启用了管理界面，可以添加： 342 | `rabbitmq-plugins enable rabbitmq_federation_management` 343 | 这样就可以在界面配置Upstream和Policy了。 344 | 345 | 注：如果在一个集群中使用federation，需要该集群每个节点都启用Federation插件 346 | 347 | 注：更多插件请见：[插件](http://www.rabbitmq.com/plugins.html) 348 | -------------------------------------------------------------------------------- /dockerfiles/rabbitmq.config: -------------------------------------------------------------------------------- 1 | %% -*- mode: erlang -*- 2 | %% ---------------------------------------------------------------------------- 3 | %% RabbitMQ Sample Configuration File. 4 | %% 5 | %% See http://www.rabbitmq.com/configure.html for details. 6 | %% ---------------------------------------------------------------------------- 7 | [ 8 | {rabbit, 9 | [%% 10 | %% Network Connectivity 11 | %% ==================== 12 | %% 13 | 14 | %% By default, RabbitMQ will listen on all interfaces, using 15 | %% the standard (reserved) AMQP port. 16 | %% 17 | %% {tcp_listeners, [5672]}, 18 | 19 | %% To listen on a specific interface, provide a tuple of {IpAddress, Port}. 20 | %% For example, to listen only on localhost for both IPv4 and IPv6: 21 | %% 22 | %% {tcp_listeners, [{"127.0.0.1", 5672}, 23 | %% {"::1", 5672}]}, 24 | 25 | %% SSL listeners are configured in the same fashion as TCP listeners, 26 | %% including the option to control the choice of interface. 27 | %% 28 | %% {ssl_listeners, [5671]}, 29 | 30 | %% Maximum time for AMQP 0-8/0-9/0-9-1 handshake (after socket connection 31 | %% and SSL handshake), in milliseconds. 32 | %% 33 | %% {handshake_timeout, 10000}, 34 | 35 | %% Log levels (currently just used for connection logging). 36 | %% One of 'debug', 'info', 'warning', 'error' or 'none', in decreasing 37 | %% order of verbosity. Defaults to 'info'. 38 | %% 39 | %% {log_levels, [{connection, info}, {channel, info}]}, 40 | 41 | %% Set to 'true' to perform reverse DNS lookups when accepting a 42 | %% connection. Hostnames will then be shown instead of IP addresses 43 | %% in rabbitmqctl and the management plugin. 44 | %% 45 | %% {reverse_dns_lookups, true}, 46 | 47 | %% 48 | %% Security / AAA 49 | %% ============== 50 | %% 51 | 52 | %% The default "guest" user is only permitted to access the server 53 | %% via a loopback interface (e.g. localhost). 54 | %% {loopback_users, [<<"guest">>]}, 55 | %% 56 | %% Uncomment the following line if you want to allow access to the 57 | %% guest user from anywhere on the network. 58 | %% {loopback_users, []}, 59 | {loopback_users, []}, 60 | 61 | %% Configuring SSL. 62 | %% See http://www.rabbitmq.com/ssl.html for full documentation. 63 | %% 64 | %% {ssl_options, [{cacertfile, "/path/to/testca/cacert.pem"}, 65 | %% {certfile, "/path/to/server/cert.pem"}, 66 | %% {keyfile, "/path/to/server/key.pem"}, 67 | %% {verify, verify_peer}, 68 | %% {fail_if_no_peer_cert, false}]}, 69 | 70 | %% Choose the available SASL mechanism(s) to expose. 71 | %% The two default (built in) mechanisms are 'PLAIN' and 72 | %% 'AMQPLAIN'. Additional mechanisms can be added via 73 | %% plugins. 74 | %% 75 | %% See http://www.rabbitmq.com/authentication.html for more details. 76 | %% 77 | %% {auth_mechanisms, ['PLAIN', 'AMQPLAIN']}, 78 | 79 | %% Select an authentication database to use. RabbitMQ comes bundled 80 | %% with a built-in auth-database, based on mnesia. 81 | %% 82 | %% {auth_backends, [rabbit_auth_backend_internal]}, 83 | 84 | %% Configurations supporting the rabbitmq_auth_mechanism_ssl and 85 | %% rabbitmq_auth_backend_ldap plugins. 86 | %% 87 | %% NB: These options require that the relevant plugin is enabled. 88 | %% See http://www.rabbitmq.com/plugins.html for further details. 89 | 90 | %% The RabbitMQ-auth-mechanism-ssl plugin makes it possible to 91 | %% authenticate a user based on the client's SSL certificate. 92 | %% 93 | %% To use auth-mechanism-ssl, add to or replace the auth_mechanisms 94 | %% list with the entry 'EXTERNAL'. 95 | %% 96 | %% {auth_mechanisms, ['EXTERNAL']}, 97 | 98 | %% The rabbitmq_auth_backend_ldap plugin allows the broker to 99 | %% perform authentication and authorisation by deferring to an 100 | %% external LDAP server. 101 | %% 102 | %% For more information about configuring the LDAP backend, see 103 | %% http://www.rabbitmq.com/ldap.html. 104 | %% 105 | %% Enable the LDAP auth backend by adding to or replacing the 106 | %% auth_backends entry: 107 | %% 108 | %% {auth_backends, [rabbit_auth_backend_ldap]}, 109 | 110 | %% This pertains to both the rabbitmq_auth_mechanism_ssl plugin and 111 | %% STOMP ssl_cert_login configurations. See the rabbitmq_stomp 112 | %% configuration section later in this file and the README in 113 | %% https://github.com/rabbitmq/rabbitmq-auth-mechanism-ssl for further 114 | %% details. 115 | %% 116 | %% To use the SSL cert's CN instead of its DN as the username 117 | %% 118 | %% {ssl_cert_login_from, common_name}, 119 | 120 | %% SSL handshake timeout, in milliseconds. 121 | %% 122 | %% {ssl_handshake_timeout, 5000}, 123 | 124 | %% 125 | %% Default User / VHost 126 | %% ==================== 127 | %% 128 | 129 | %% On first start RabbitMQ will create a vhost and a user. These 130 | %% config items control what gets created. See 131 | %% http://www.rabbitmq.com/access-control.html for further 132 | %% information about vhosts and access control. 133 | %% 134 | %% {default_vhost, <<"/">>}, 135 | {default_user, <<"admin">>}, 136 | {default_pass, <<"admin">>}, 137 | {default_permissions, [<<".*">>, <<".*">>, <<".*">>]}, 138 | 139 | %% Tags for default user 140 | %% 141 | %% For more details about tags, see the documentation for the 142 | %% Management Plugin at http://www.rabbitmq.com/management.html. 143 | %% 144 | {default_user_tags, [administrator]}, 145 | 146 | %% 147 | %% Additional network and protocol related configuration 148 | %% ===================================================== 149 | %% 150 | 151 | %% Set the default AMQP heartbeat delay (in seconds). 152 | %% 153 | %% {heartbeat, 600}, 154 | 155 | %% Set the max permissible size of an AMQP frame (in bytes). 156 | %% 157 | %% {frame_max, 131072}, 158 | 159 | %% Set the max permissible number of channels per connection. 160 | %% 0 means "no limit". 161 | %% 162 | %% {channel_max, 128}, 163 | 164 | %% Customising Socket Options. 165 | %% 166 | %% See (http://www.erlang.org/doc/man/inet.html#setopts-2) for 167 | %% further documentation. 168 | %% 169 | %% {tcp_listen_options, [binary, 170 | %% {packet, raw}, 171 | %% {reuseaddr, true}, 172 | %% {backlog, 128}, 173 | %% {nodelay, true}, 174 | %% {exit_on_close, false}]}, 175 | 176 | %% 177 | %% Resource Limits & Flow Control 178 | %% ============================== 179 | %% 180 | %% See http://www.rabbitmq.com/memory.html for full details. 181 | 182 | %% Memory-based Flow Control threshold. 183 | %% 184 | %% {vm_memory_high_watermark, 0.4}, 185 | {vm_memory_high_watermark, 0.2}, 186 | 187 | %% Fraction of the high watermark limit at which queues start to 188 | %% page message out to disc in order to free up memory. 189 | %% 190 | %% Values greater than 0.9 can be dangerous and should be used carefully. 191 | %% 192 | %% {vm_memory_high_watermark_paging_ratio, 0.5}, 193 | 194 | %% Set disk free limit (in bytes). Once free disk space reaches this 195 | %% lower bound, a disk alarm will be set - see the documentation 196 | %% listed above for more details. 197 | %% 198 | %% {disk_free_limit, 50000000}, 199 | {disk_free_limit, 1000000000} 200 | 201 | %% Alternatively, we can set a limit relative to total available RAM. 202 | %% 203 | %% Values lower than 1.0 can be dangerous and should be used carefully. 204 | %% {disk_free_limit, {mem_relative, 2.0}}, 205 | 206 | %% 207 | %% Misc/Advanced Options 208 | %% ===================== 209 | %% 210 | %% NB: Change these only if you understand what you are doing! 211 | %% 212 | 213 | %% To announce custom properties to clients on connection: 214 | %% 215 | %% {server_properties, []}, 216 | 217 | %% How to respond to cluster partitions. 218 | %% See http://www.rabbitmq.com/partitions.html for further details. 219 | %% 220 | %% {cluster_partition_handling, ignore}, 221 | 222 | %% Make clustering happen *automatically* at startup - only applied 223 | %% to nodes that have just been reset or started for the first time. 224 | %% See http://www.rabbitmq.com/clustering.html#auto-config for 225 | %% further details. 226 | %% 227 | %% {cluster_nodes, {['rabbit@my.host.com'], disc}}, 228 | %% {cluster_nodes, {['rabbit@ip-10-20-22-11','rabbit@ip-10-20-20-12'], disc}} 229 | %% {cluster_nodes, {['rabbit@boot2docker','rabbit2@boot2docker'], disc}} 230 | 231 | %% Interval (in milliseconds) at which we send keepalive messages 232 | %% to other cluster members. Note that this is not the same thing 233 | %% as net_ticktime; missed keepalive messages will not cause nodes 234 | %% to be considered down. 235 | %% 236 | %% {cluster_keepalive_interval, 10000}, 237 | 238 | %% Set (internal) statistics collection granularity. 239 | %% 240 | %% {collect_statistics, none}, 241 | 242 | %% Statistics collection interval (in milliseconds). 243 | %% 244 | %% {collect_statistics_interval, 5000}, 245 | 246 | %% Explicitly enable/disable hipe compilation. 247 | %% 248 | %% {hipe_compile, true}, 249 | 250 | %% Timeout used when waiting for Mnesia tables in a cluster to 251 | %% become available. 252 | %% 253 | %% {mnesia_table_loading_timeout, 30000}, 254 | 255 | %% Size in bytes below which to embed messages in the queue index. See 256 | %% http://www.rabbitmq.com/persistence-conf.html 257 | %% 258 | %% {queue_index_embed_msgs_below, 4096} 259 | 260 | ]}, 261 | 262 | %% ---------------------------------------------------------------------------- 263 | %% Advanced Erlang Networking/Clustering Options. 264 | %% 265 | %% See http://www.rabbitmq.com/clustering.html for details 266 | %% ---------------------------------------------------------------------------- 267 | {kernel, 268 | [%% Sets the net_kernel tick time. 269 | %% Please see http://erlang.org/doc/man/kernel_app.html and 270 | %% http://www.rabbitmq.com/nettick.html for further details. 271 | %% 272 | %% {net_ticktime, 60} 273 | ]}, 274 | 275 | %% ---------------------------------------------------------------------------- 276 | %% RabbitMQ Management Plugin 277 | %% 278 | %% See http://www.rabbitmq.com/management.html for details 279 | %% ---------------------------------------------------------------------------- 280 | 281 | {rabbitmq_management, 282 | [%% Pre-Load schema definitions from the following JSON file. See 283 | %% http://www.rabbitmq.com/management.html#load-definitions 284 | %% 285 | %% {load_definitions, "/path/to/schema.json"}, 286 | 287 | %% Log all requests to the management HTTP API to a file. 288 | %% 289 | %% {http_log_dir, "/path/to/access.log"}, 290 | 291 | %% Change the port on which the HTTP listener listens, 292 | %% specifying an interface for the web server to bind to. 293 | %% Also set the listener to use SSL and provide SSL options. 294 | %% 295 | %% {listener, [{port, 12345}, 296 | %% {ip, "127.0.0.1"}, 297 | %% {ssl, true}, 298 | %% {ssl_opts, [{cacertfile, "/path/to/cacert.pem"}, 299 | %% {certfile, "/path/to/cert.pem"}, 300 | %% {keyfile, "/path/to/key.pem"}]}]}, 301 | 302 | %% One of 'basic', 'detailed' or 'none'. See 303 | %% http://www.rabbitmq.com/management.html#fine-stats for more details. 304 | %% {rates_mode, basic}, 305 | 306 | %% Configure how long aggregated data (such as message rates and queue 307 | %% lengths) is retained. Please read the plugin's documentation in 308 | %% http://www.rabbitmq.com/management.html#configuration for more 309 | %% details. 310 | %% 311 | %% {sample_retention_policies, 312 | %% [{global, [{60, 5}, {3600, 60}, {86400, 1200}]}, 313 | %% {basic, [{60, 5}, {3600, 60}]}, 314 | %% {detailed, [{10, 5}]}]} 315 | ]}, 316 | 317 | %% ---------------------------------------------------------------------------- 318 | %% RabbitMQ Shovel Plugin 319 | %% 320 | %% See http://www.rabbitmq.com/shovel.html for details 321 | %% ---------------------------------------------------------------------------- 322 | 323 | {rabbitmq_shovel, 324 | [{shovels, 325 | [%% A named shovel worker. 326 | %% {my_first_shovel, 327 | %% [ 328 | 329 | %% List the source broker(s) from which to consume. 330 | %% 331 | %% {sources, 332 | %% [%% URI(s) and pre-declarations for all source broker(s). 333 | %% {brokers, ["amqp://user:password@host.domain/my_vhost"]}, 334 | %% {declarations, []} 335 | %% ]}, 336 | 337 | %% List the destination broker(s) to publish to. 338 | %% {destinations, 339 | %% [%% A singular version of the 'brokers' element. 340 | %% {broker, "amqp://"}, 341 | %% {declarations, []} 342 | %% ]}, 343 | 344 | %% Name of the queue to shovel messages from. 345 | %% 346 | %% {queue, <<"your-queue-name-goes-here">>}, 347 | 348 | %% Optional prefetch count. 349 | %% 350 | %% {prefetch_count, 10}, 351 | 352 | %% when to acknowledge messages: 353 | %% - no_ack: never (auto) 354 | %% - on_publish: after each message is republished 355 | %% - on_confirm: when the destination broker confirms receipt 356 | %% 357 | %% {ack_mode, on_confirm}, 358 | 359 | %% Overwrite fields of the outbound basic.publish. 360 | %% 361 | %% {publish_fields, [{exchange, <<"my_exchange">>}, 362 | %% {routing_key, <<"from_shovel">>}]}, 363 | 364 | %% Static list of basic.properties to set on re-publication. 365 | %% 366 | %% {publish_properties, [{delivery_mode, 2}]}, 367 | 368 | %% The number of seconds to wait before attempting to 369 | %% reconnect in the event of a connection failure. 370 | %% 371 | %% {reconnect_delay, 2.5} 372 | 373 | %% ]} %% End of my_first_shovel 374 | ]} 375 | %% Rather than specifying some values per-shovel, you can specify 376 | %% them for all shovels here. 377 | %% 378 | %% {defaults, [{prefetch_count, 0}, 379 | %% {ack_mode, on_confirm}, 380 | %% {publish_fields, []}, 381 | %% {publish_properties, [{delivery_mode, 2}]}, 382 | %% {reconnect_delay, 2.5}]} 383 | ]}, 384 | 385 | %% ---------------------------------------------------------------------------- 386 | %% RabbitMQ Stomp Adapter 387 | %% 388 | %% See http://www.rabbitmq.com/stomp.html for details 389 | %% ---------------------------------------------------------------------------- 390 | 391 | {rabbitmq_stomp, 392 | [%% Network Configuration - the format is generally the same as for the broker 393 | 394 | %% Listen only on localhost (ipv4 & ipv6) on a specific port. 395 | %% {tcp_listeners, [{"127.0.0.1", 61613}, 396 | %% {"::1", 61613}]}, 397 | 398 | %% Listen for SSL connections on a specific port. 399 | %% {ssl_listeners, [61614]}, 400 | 401 | %% Additional SSL options 402 | 403 | %% Extract a name from the client's certificate when using SSL. 404 | %% 405 | %% {ssl_cert_login, true}, 406 | 407 | %% Set a default user name and password. This is used as the default login 408 | %% whenever a CONNECT frame omits the login and passcode headers. 409 | %% 410 | %% Please note that setting this will allow clients to connect without 411 | %% authenticating! 412 | %% 413 | %% {default_user, [{login, "guest"}, 414 | %% {passcode, "guest"}]}, 415 | 416 | %% If a default user is configured, or you have configured use SSL client 417 | %% certificate based authentication, you can choose to allow clients to 418 | %% omit the CONNECT frame entirely. If set to true, the client is 419 | %% automatically connected as the default user or user supplied in the 420 | %% SSL certificate whenever the first frame sent on a session is not a 421 | %% CONNECT frame. 422 | %% 423 | %% {implicit_connect, true} 424 | ]}, 425 | 426 | %% ---------------------------------------------------------------------------- 427 | %% RabbitMQ MQTT Adapter 428 | %% 429 | %% See https://github.com/rabbitmq/rabbitmq-mqtt/blob/stable/README.md 430 | %% for details 431 | %% ---------------------------------------------------------------------------- 432 | 433 | {rabbitmq_mqtt, 434 | [%% Set the default user name and password. Will be used as the default login 435 | %% if a connecting client provides no other login details. 436 | %% 437 | %% Please note that setting this will allow clients to connect without 438 | %% authenticating! 439 | %% 440 | %% {default_user, <<"guest">>}, 441 | %% {default_pass, <<"guest">>}, 442 | 443 | %% Enable anonymous access. If this is set to false, clients MUST provide 444 | %% login information in order to connect. See the default_user/default_pass 445 | %% configuration elements for managing logins without authentication. 446 | %% 447 | %% {allow_anonymous, true}, 448 | 449 | %% If you have multiple chosts, specify the one to which the 450 | %% adapter connects. 451 | %% 452 | %% {vhost, <<"/">>}, 453 | 454 | %% Specify the exchange to which messages from MQTT clients are published. 455 | %% 456 | %% {exchange, <<"amq.topic">>}, 457 | 458 | %% Specify TTL (time to live) to control the lifetime of non-clean sessions. 459 | %% 460 | %% {subscription_ttl, 1800000}, 461 | 462 | %% Set the prefetch count (governing the maximum number of unacknowledged 463 | %% messages that will be delivered). 464 | %% 465 | %% {prefetch, 10}, 466 | 467 | %% TCP/SSL Configuration (as per the broker configuration). 468 | %% 469 | %% {tcp_listeners, [1883]}, 470 | %% {ssl_listeners, []}, 471 | 472 | %% TCP/Socket options (as per the broker configuration). 473 | %% 474 | %% {tcp_listen_options, [binary, 475 | %% {packet, raw}, 476 | %% {reuseaddr, true}, 477 | %% {backlog, 128}, 478 | %% {nodelay, true}]} 479 | ]}, 480 | 481 | %% ---------------------------------------------------------------------------- 482 | %% RabbitMQ AMQP 1.0 Support 483 | %% 484 | %% See https://github.com/rabbitmq/rabbitmq-amqp1.0/blob/stable/README.md 485 | %% for details 486 | %% ---------------------------------------------------------------------------- 487 | 488 | {rabbitmq_amqp1_0, 489 | [%% Connections that are not authenticated with SASL will connect as this 490 | %% account. See the README for more information. 491 | %% 492 | %% Please note that setting this will allow clients to connect without 493 | %% authenticating! 494 | %% 495 | %% {default_user, "guest"}, 496 | 497 | %% Enable protocol strict mode. See the README for more information. 498 | %% 499 | %% {protocol_strict_mode, false} 500 | ]}, 501 | 502 | %% ---------------------------------------------------------------------------- 503 | %% RabbitMQ LDAP Plugin 504 | %% 505 | %% See http://www.rabbitmq.com/ldap.html for details. 506 | %% 507 | %% ---------------------------------------------------------------------------- 508 | 509 | {rabbitmq_auth_backend_ldap, 510 | [%% 511 | %% Connecting to the LDAP server(s) 512 | %% ================================ 513 | %% 514 | 515 | %% Specify servers to bind to. You *must* set this in order for the plugin 516 | %% to work properly. 517 | %% 518 | %% {servers, ["your-server-name-goes-here"]}, 519 | 520 | %% Connect to the LDAP server using SSL 521 | %% 522 | %% {use_ssl, false}, 523 | 524 | %% Specify the LDAP port to connect to 525 | %% 526 | %% {port, 389}, 527 | 528 | %% LDAP connection timeout, in milliseconds or 'infinity' 529 | %% 530 | %% {timeout, infinity}, 531 | 532 | %% Enable logging of LDAP queries. 533 | %% One of 534 | %% - false (no logging is performed) 535 | %% - true (verbose logging of the logic used by the plugin) 536 | %% - network (as true, but additionally logs LDAP network traffic) 537 | %% 538 | %% Defaults to false. 539 | %% 540 | %% {log, false}, 541 | 542 | %% 543 | %% Authentication 544 | %% ============== 545 | %% 546 | 547 | %% Pattern to convert the username given through AMQP to a DN before 548 | %% binding 549 | %% 550 | %% {user_dn_pattern, "cn=${username},ou=People,dc=example,dc=com"}, 551 | 552 | %% Alternatively, you can convert a username to a Distinguished 553 | %% Name via an LDAP lookup after binding. See the documentation for 554 | %% full details. 555 | 556 | %% When converting a username to a dn via a lookup, set these to 557 | %% the name of the attribute that represents the user name, and the 558 | %% base DN for the lookup query. 559 | %% 560 | %% {dn_lookup_attribute, "userPrincipalName"}, 561 | %% {dn_lookup_base, "DC=gopivotal,DC=com"}, 562 | 563 | %% Controls how to bind for authorisation queries and also to 564 | %% retrieve the details of users logging in without presenting a 565 | %% password (e.g., SASL EXTERNAL). 566 | %% One of 567 | %% - as_user (to bind as the authenticated user - requires a password) 568 | %% - anon (to bind anonymously) 569 | %% - {UserDN, Password} (to bind with a specified user name and password) 570 | %% 571 | %% Defaults to 'as_user'. 572 | %% 573 | %% {other_bind, as_user}, 574 | 575 | %% 576 | %% Authorisation 577 | %% ============= 578 | %% 579 | 580 | %% The LDAP plugin can perform a variety of queries against your 581 | %% LDAP server to determine questions of authorisation. See 582 | %% http://www.rabbitmq.com/ldap.html#authorisation for more 583 | %% information. 584 | 585 | %% Set the query to use when determining vhost access 586 | %% 587 | %% {vhost_access_query, {in_group, 588 | %% "ou=${vhost}-users,ou=vhosts,dc=example,dc=com"}}, 589 | 590 | %% Set the query to use when determining resource (e.g., queue) access 591 | %% 592 | %% {resource_access_query, {constant, true}}, 593 | 594 | %% Set queries to determine which tags a user has 595 | %% 596 | %% {tag_queries, []} 597 | ]} 598 | ]. 599 | --------------------------------------------------------------------------------