文章目录
概述
传统的消息传递模式有2种:队列( queue) 和(publish-subscribe)
- queue模式:多个consumer从服务器中读取数据,消息只会到达一个consumer
- publish-subscribe模式:消息会被广播给所有的consumer
Kafka基于这2种模式提供了一种consumer的抽象概念: consumer group
- queue模式:所有的consumer都位于同一个consumer group 下。
- publish-subscribe模式:所有的consumer都有着自己唯一的consumer group
说明: 由2个broker组成的kafka集群,某个主题总共有4个partition(P0-P3),分别位于不同的broker上。这个集群由2个Consumer Group消费, A有2个consumer instances ,B有4个。
通常一个topic会有几个consumer group,每个consumer group都是一个逻辑上的订阅者( logicalsubscriber )。每个consumer group由多个consumer instance组成,从而达到可扩展和容灾的功能。
广播模式的应用 ----> 应用里缓存了数据字典等配置表在内存中,可以通过 Kafka 广播消费,实现每个应用节点都消费消息,刷新本地内存的缓存。
Code
POM依赖
<dependencies>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-web</artifactId>
</dependency>
<!-- 引入 Spring-Kafka 依赖 -->
<dependency>
<groupId>org.springframework.kafka</groupId>
<artifactId>spring-kafka</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-test</artifactId>
<scope>test</scope>
</dependency>
<dependency>
<groupId>junit</groupId>
<artifactId>junit</artifactId>
<scope>test</scope>
</dependency>
</dependencies>
配置文件
spring:
# Kafka 配置项,对应 KafkaProperties 配置类
kafka:
bootstrap-servers: 192.168.126.140:9092 # 指定 Kafka Broker 地址,可以设置多个,以逗号分隔
# Kafka Producer 配置项
producer:
acks: 1 # 0-不应答。1-leader 应答。all-所有 leader 和 follower 应答。
retries: 3 # 发送失败时,重试发送的次数
key-serializer: org.apache.kafka.common.serialization.StringSerializer # 消息的 key 的序列化
value-serializer: org.springframework.kafka.support.serializer.JsonSerializer # 消息的 value 的序列化
# Kafka Consumer 配置项
consumer:
auto-offset-reset: latest # 在广播订阅下,一般情况下,无需消费历史的消息,而是从订阅的 Topic 的队列的尾部开始消费即可
key-deserializer: org.apache.kafka.common.serialization.StringDeserializer
value-deserializer: org.springframework.kafka.support.serializer.JsonDeserializer
properties:
spring:
json:
trusted:
packages: com.artisan.springkafka.domain
# Kafka Consumer Listener 监听器配置
listener:
missing-topics-fatal: false # 消费监听接口监听的主题不存在时,默认会报错。所以通过设置为 false ,解决报错
logging:
level:
org:
springframework:
kafka: ERROR # spring-kafka
apache:
kafka: ERROR # kafka
auto-offset-reset: latest
广播模式,一般情况下,无需消费历史的消息,从订阅的 Topic 的队列的尾部开始消费即可
生产者
package com.artisan.springkafka.producer;
import com.artisan.springkafka.constants.TOPIC;
import com.artisan.springkafka.domain.MessageMock;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.kafka.core.KafkaTemplate;
import org.springframework.kafka.support.SendResult;
import org.springframework.stereotype.Component;
import org.springframework.util.concurrent.ListenableFuture;
import java.util.Random;
import java.util.concurrent.ExecutionException;
/**
* @author 小工匠
* @version 1.0
* @description: TODO
* @date 2021/2/17 22:25
* @mark: show me the code , change the world
*/
@Component
public class ArtisanProducerMock {
@Autowired
private KafkaTemplate<Object,Object> kafkaTemplate ;
/**
* 异步发送
* @return
* @throws ExecutionException
* @throws InterruptedException
*/
public ListenableFuture<SendResult<Object, Object>> sendMsgASync() throws ExecutionException, InterruptedException {
// 模拟发送的消息
Integer id = new Random().nextInt(100);
MessageMock messageMock = new MessageMock(id,"messageSendByAsync-" + id);
// 异步发送消息
ListenableFuture<SendResult<Object, Object>> result = kafkaTemplate.send(TOPIC.TOPIC, messageMock);
return result ;
}
}
消费者
package com.artisan.springkafka.consumer;
import com.artisan.springkafka.domain.MessageMock;
import com.artisan.springkafka.constants.TOPIC;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.kafka.annotation.KafkaListener;
import org.springframework.stereotype.Component;
/**
* @author 小工匠
* @version 1.0
* @description: TODO
* @date 2021/2/17 22:33
* @mark: show me the code , change the world
*/
@Component
public class ArtisanCosumerMockDiffConsumeGroup {
private Logger logger = LoggerFactory.getLogger(getClass());
private static final String CONSUMER_GROUP_PREFIX = "MOCK-B" ;
@KafkaListener(topics = TOPIC.TOPIC ,groupId = CONSUMER_GROUP_PREFIX + TOPIC.TOPIC + "-" + "#{T(java.util.UUID).randomUUID()})")
public void onMessage(MessageMock messageMock){
logger.info("【接受到消息][线程:{} 消息内容:{}]", Thread.currentThread().getName(), messageMock);
}
}
注意: groupId 通过 Spring EL 表达式,在每个消费者分组的名字上配合 UUID 生成其后缀。这样,就能保证每个项目启动的消费者分组不同,从而达到广播消费的目的。
单元测试
package com.artisan.springkafka.produceTest;
import com.artisan.springkafka.SpringkafkaApplication;
import com.artisan.springkafka.producer.ArtisanProducerMock;
import org.junit.Test;
import org.junit.runner.RunWith;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.boot.test.context.SpringBootTest;
import org.springframework.kafka.support.SendResult;
import org.springframework.test.context.junit4.SpringRunner;
import org.springframework.util.concurrent.ListenableFuture;
import org.springframework.util.concurrent.ListenableFutureCallback;
import java.util.concurrent.CountDownLatch;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeUnit;
/**
* @author 小工匠
* * @version 1.0
* @description: TODO
* @date 2021/2/17 22:40
* @mark: show me the code , change the world
*/
@RunWith(SpringRunner.class)
@SpringBootTest(classes = SpringkafkaApplication.class)
public class ProduceMockTest {
private Logger logger = LoggerFactory.getLogger(getClass());
@Autowired
private ArtisanProducerMock artisanProducerMock;
@Test
public void testAsynSend() throws ExecutionException, InterruptedException {
logger.info("开始发送 测试广播模式");
artisanProducerMock.sendMsgASync().addCallback(new ListenableFutureCallback<SendResult<Object, Object>>() {
@Override
public void onFailure(Throwable throwable) {
logger.info(" 发送异常{}]]", throwable);
}
@Override
public void onSuccess(SendResult<Object, Object> objectObjectSendResult) {
logger.info("回调结果 Result = topic:[{}] , partition:[{}], offset:[{}]",
objectObjectSendResult.getRecordMetadata().topic(),
objectObjectSendResult.getRecordMetadata().partition(),
objectObjectSendResult.getRecordMetadata().offset());
}
});
// 阻塞等待,保证消费
new CountDownLatch(1).await();
}
}
启动多次单元测试, 观察消息的接受情况
测速结果
可以看到不消费组下的 消费者(目前是一个消费组下一个消费者) 均收到了 这条消息,这就是广播模式
源码地址
https://github.com/yangshangwei/boot2/tree/master/springkafkaBroadCast