在发送消息时,发送方可以自定义消息的用户属性,消费者可以利用SQL92的WHERE子句语法实现消息过滤。
相比Tag过滤,消息过滤使用更加灵活,也更容易被程序猿接受,但相较Tag过滤执行效率较低。
消息生产者
消息发送方和标准发送有两点变化:
● 可以不设置消息的Tag与Key,转而使用用户自定义属性,这里实现了source与id两个自定义属性的赋值
● 利用message.putUserProperty为用户赋予自定义属性
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 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
| @Slf4j public class SfProducer { public static void main(String[] args) { DefaultMQProducer producer = new DefaultMQProducer("sf-producer-group"); producer.setNamesrvAddr("192.168.31.103:9876"); try { producer.start(); for(Integer i = 0 ; i < 10 ; i++) { Thread.sleep(1000); Integer rnd = new Random().nextInt(10); String source = ""; switch (rnd % 3){ case 0: source = "jd"; break; case 1: source = "tmall"; break; case 2: source = "taobao"; break; } String data = "第" + i + "条消息数据"; Message message = new Message("sf-sample-data", data.getBytes()); message.putUserProperty("id" , i.toString()); message.putUserProperty("source", source); SendResult result = producer.send(message); log.info("id:{},source:{},data:{}" ,i.toString(), source,data); } }catch (Exception e){ e.printStackTrace(); }finally { try { producer.shutdown(); log.info("连接已关闭"); }catch (Exception e){ e.printStackTrace(); } } } }
|
运行结果
1 2 3 4 5 6 7 8 9 10
| 09:34:52.771 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:0,source:jd,data:第0条消息数据 09:34:53.788 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:1,source:jd,data:第1条消息数据 09:34:54.801 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:2,source:tmall,data:第2条消息数据 09:34:55.818 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:3,source:tmall,data:第3条消息数据 09:34:56.836 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:4,source:jd,data:第4条消息数据 09:34:57.850 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:5,source:taobao,data:第5条消息数据 09:34:58.865 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:6,source:taobao,data:第6条消息数据 09:34:59.880 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:7,source:jd,data:第7条消息数据 09:35:00.896 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:8,source:tmall,data:第8条消息数据 09:35:01.911 [main] INFO com.lixiang.rocketmq.sqlfilter.SfProducer - id:9,source:taobao,data:第9条消息数据
|
消息消费者
默认RocketMQ并未开启自定义属性SQL过滤的选项,需要在配置文件中额外开启,如下所示:
master.conf:Master节点配置文件追加下面选型:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
| #开启自定义属性SQL过滤 enablePropertyFilter=true 完整如下: brokerClusterName=DefaultCluster brokerName=broker-a brokerId=0 deleteWhen=04 fileReservedTime=48 brokerRole=SYNC_MASTER flushDiskType=SYNC_FLUSH namesrvAddr=192.168.31.103:9876 autoCreateTopicEnable=true #开启自定义属性SQL过滤 enablePropertyFilter=true slave.conf:Slave节点也要追加该配置项,别忘记
|
京东消费者
京东消费者负责消费source=’jd’的数据,和标准消费者最大的不同便是在subscribe方法第二个参数不再是Tag,而改为MessageSelector.bySql方法,利用WHERE子句写法对自定义属性实现过滤,代码如下:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
| @Slf4j public class SfJDConsumer { public static void main(String[] args) throws Exception { DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("sf-jd-consumer-group"); consumer.setNamesrvAddr("192.168.31.103:9876"); consumer.setMessageModel(MessageModel.CLUSTERING); consumer.subscribe("sf-sample-data", MessageSelector.bySql("source='jd'")); consumer.registerMessageListener(new MessageListenerConcurrently() { @Override public ConsumeConcurrentlyStatus consumeMessage(List<MessageExt> list, ConsumeConcurrentlyContext consumeConcurrentlyContext) { list.forEach(msg->{ log.info("id:{},source:{},data:{}" ,msg.getUserProperty("id"),msg.getUserProperty("source"), new String(msg.getBody()));}); return ConsumeConcurrentlyStatus.CONSUME_SUCCESS; } }); consumer.start(); log.info("集群消费者启动成功,正在监听新消息"); } }
|
运行结果
1 2 3 4
| 09:34:52.770 [ConsumeMessageThread_2] INFO com.lixiang.rocketmq.sqlfilter.SfJDConsumer - id:0,source:jd,data:第0条消息数据 09:34:53.788 [ConsumeMessageThread_3] INFO com.lixiang.rocketmq.sqlfilter.SfJDConsumer - id:1,source:jd,data:第1条消息数据 09:34:56.836 [ConsumeMessageThread_4] INFO com.lixiang.rocketmq.sqlfilter.SfJDConsumer - id:4,source:jd,data:第4条消息数据 09:34:59.880 [ConsumeMessageThread_5] INFO com.lixiang.rocketmq.sqlfilter.SfJDConsumer - id:7,source:jd,data:第7条消息数据
|
阿里消费者
京东消费者负责消费天与猫淘宝的数据,与京东消费者最明显的区别是:
● 因为业务范围不同,消费者组不一样;
● bySQL的要获取多个数值,可用下面语法
● source in (‘tmall’,’taobao’)
● source = ‘tmall’ or source = ‘taobao’
1 2 3 4 5
| ... DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("sf-ali-consumer-group"); ... consumer.subscribe("sf-sample-data", MessageSelector.bySql("source in ('tmall' ,'taobao')"));
|
运行结果
1 2 3 4 5 6
| 09:34:54.803 [ConsumeMessageThread_14] INFO com.lixiang.rocketmq.sqlfilter.SfAliConsumer - id:2,source:tmall,data:第2条消息数据 09:34:55.819 [ConsumeMessageThread_15] INFO com.lixiang.rocketmq.sqlfilter.SfAliConsumer - id:3,source:tmall,data:第3条消息数据 09:34:57.850 [ConsumeMessageThread_16] INFO com.lixiang.rocketmq.sqlfilter.SfAliConsumer - id:5,source:taobao,data:第5条消息数据 09:34:58.865 [ConsumeMessageThread_17] INFO com.lixiang.rocketmq.sqlfilter.SfAliConsumer - id:6,source:taobao,data:第6条消息数据 09:35:00.896 [ConsumeMessageThread_18] INFO com.lixiang.rocketmq.sqlfilter.SfAliConsumer - id:8,source:tmall,data:第8条消息数据 09:35:01.911 [ConsumeMessageThread_19] INFO com.lixiang.rocketmq.sqlfilter.SfAliConsumer - id:9,source:taobao,data:第9条消息数据
|