概述

一个partition同一个时刻在一个consumer group中只能有一个consumer instance在消费，从而保证消费顺序。

consumer group中的consumer instance的数量不能比一个Topic中的partition的数量多，否则，多出来的consumer消费不到消息。

Kafka只在partition的范围内保证消息消费的局部顺序性，不能在同一个topic中的多个partition中保证总的消费顺序性。

如果有在总体上保证消费顺序的需求，那么我们可以通过将topic的partition数量设置为1，将consumer group中的consumer instance数量也设置为1，但是这样会影响性能，所以kafka的顺序消费很少用。

有的时候，我们需要根据某个主键来实现消费的有序性，那我们只需要考虑将 Producer 将相关联的消息发送到 Topic 下的相同的 Partition 即可。 Kafka Producer 发送消息使用分区策略的话，根据 key 的哈希值取模来获取到其在 Topic 下对应的 Partition 。 所以我们发送消息的时候指定key即可

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: earliest # 设置消费者分组最初的消费进度为 earliest
      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

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 ;


  
    /**
     * 同步发送
     * @return
     * @throws ExecutionException
     * @throws InterruptedException
     */
    public SendResult sendMsgSync() throws ExecutionException, InterruptedException {
        // 模拟发送的消息
        int id = 6687421;
        MessageMock messageMock = new MessageMock(id,"artisanTestMessage-" + id);


        // 测试顺序消息 以 id 为key  (因为使用 String 的方式序列化 key ，所以需要将 id 转换成 String)
        return kafkaTemplate.send(TOPIC.TOPIC,String.valueOf(id) ,messageMock).get();
    }

    /**
     * 同步发送
     * @return
     * @throws ExecutionException
     * @throws InterruptedException
     */
    public SendResult sendMsgSync2() throws ExecutionException, InterruptedException {
        // 模拟发送的消息
        int id = 1255447;
        MessageMock messageMock = new MessageMock(id,"artisanTestMessage-" + id);


        // 测试顺序消息 以 id 为key  (因为使用 String 的方式序列化 key ，所以需要将 id 转换成 String)
        return kafkaTemplate.send(TOPIC.TOPIC,String.valueOf(id) ,messageMock).get();
    }
}

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 ArtisanCosumerMock {


    private Logger logger = LoggerFactory.getLogger(getClass());
    private static final String CONSUMER_GROUP_PREFIX = "MOCK-A" ;

    @KafkaListener(topics = TOPIC.TOPIC ,groupId = CONSUMER_GROUP_PREFIX + TOPIC.TOPIC,
        concurrency = "2")
    public void onMessage(MessageMock messageMock){
//        logger.info("【接受到消息][线程ID:{} 消息内容：{}]", Thread.currentThread().getId(), messageMock);
    }


}

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");

        // 模拟发送多条消息
        for (int i = 0; i < 5; i++) {
            SendResult sendResult = artisanProducerMock.sendMsgSync();

            logger.info("key {} . 分区:{}",sendResult.getProducerRecord().key(),sendResult.getRecordMetadata().partition());

            SendResult sendResult1 = artisanProducerMock.sendMsgSync2();

            logger.info("key {} . 分区:{}",sendResult1.getProducerRecord().key(),sendResult1.getRecordMetadata().partition());
        }


        // 阻塞等待，保证消费
        new CountDownLatch(1).await();

    }

}

测试结果

2021-02-18 23:05:46.903  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : 开始发送
2021-02-18 23:05:47.088  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 6687421 . 分区:1
2021-02-18 23:05:47.090  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 1255447 . 分区:0
2021-02-18 23:05:47.093  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 6687421 . 分区:1
2021-02-18 23:05:47.096  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 1255447 . 分区:0
2021-02-18 23:05:47.097  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 6687421 . 分区:1
2021-02-18 23:05:47.099  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 1255447 . 分区:0
2021-02-18 23:05:47.101  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 6687421 . 分区:1
2021-02-18 23:05:47.103  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 1255447 . 分区:0
2021-02-18 23:05:47.105  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 6687421 . 分区:1
2021-02-18 23:05:47.106  INFO 24568 --- [           main] c.a.s.produceTest.ProduceMockTest        : key 1255447 . 分区:0