架构图：（网图，很通俗易懂了，就不自己画了，这里实现的是一个Producer 两个Consumer）

 

前提：已经开启zookeeper 和kafka ，具体可参考博客https://blog.csdn.net/DGH2430284817/article/details/90483089

 

步骤：

        1,：设置kafka 分区为2 ：

        修改kafka 目录config 下的文件server.properties ，修改 num.partitions=2，保存，重启kafka。

# The default number of log partitions per topic. More partitions allow greater
# parallelism for consumption, but this will also result in more files across
# the brokers.
num.partitions=2

        在修改后，生产者在发送消息到kafka 的broker 的时候就会保存到两个分区中的一个，具体是哪个要根据分区规则的写法，一般采取的都是生产者提供一个key ，对这个key 进行hashCode() 计算它的值，再对这个值除分区数取余，用这个余数来决定消息保存到哪个分区。（如现在分区是2，经过上面的计算后只会出现0和1的值，而且是散列的，两个分区的数据量会差不多）

        2：编写分区规则：

package com.kafka1;
import kafka.producer.Partitioner;
import kafka.utils.VerifiableProperties;
 
public class SimplePartitioner implements Partitioner {
	
	public SimplePartitioner () { }
	
	public SimplePartitioner (VerifiableProperties props) { }
 
	public int partition(Object key, int numPartitions) {
		int partition = 0;
		String k = (String)key;
		partition = Math.abs(k.hashCode()) % numPartitions;//根据key的hashCode和分区数numPartitions，算出所在分区，如分区数为2，返回值只会是0和1
		return partition;
	}
} 

 

        3：编写生产者类PartitionerProducer.java：

package com.kafka1;

import java.util.Properties;
import kafka.javaapi.producer.Producer;
import kafka.producer.KeyedMessage;
import kafka.producer.ProducerConfig;
import kafka.producer.Partitioner;

public class PartitionerProducer {
	public static void main(String[] args) {
		Properties props = new Properties();
		props.put("serializer.class", "kafka.serializer.StringEncoder"); 
		props.put("metadata.broker.list", "127.0.0.1:9092");//连接 kafka
		props.put("partitioner.class", "com.kafka1.SimplePartitioner"); //指定分区规则的类
		Producer<String, String> producer = new Producer<String, String>(new ProducerConfig(props));
		String topic = "dgh-test";
		Partitioner Partitioner = new SimplePartitioner();
		for (int i = 0; i <= 10; i++) {
			String k = "key" + System.currentTimeMillis(); //key+当前毫秒，形成唯一key，也可以使用时间戳
			String v = k + "：value" + i; //消息内容
			int partition = Partitioner.partition(k, 2); //计算消息key所保存的分区号，2是分区数，实际生产这里可以不用，因为要方便观看所以输出分区
			System.out.println("topic:" + topic + ", partition:" + partition + ", key:" + k + ", value:" +v);
			producer.send(new KeyedMessage<String, String>(topic, k, v));
		}
		producer.close();
	}
}

        右键运行：

topic:dgh-test, partition:1, key:key1560695392884, value:key1560695392884：value0
topic:dgh-test, partition:1, key:key1560695393122, value:key1560695393122：value1
topic:dgh-test, partition:1, key:key1560695393124, value:key1560695393124：value2
topic:dgh-test, partition:0, key:key1560695393127, value:key1560695393127：value3
topic:dgh-test, partition:1, key:key1560695393128, value:key1560695393128：value4
topic:dgh-test, partition:0, key:key1560695393129, value:key1560695393129：value5
topic:dgh-test, partition:0, key:key1560695393130, value:key1560695393130：value6
topic:dgh-test, partition:0, key:key1560695393132, value:key1560695393132：value7
topic:dgh-test, partition:1, key:key1560695393133, value:key1560695393133：value8
topic:dgh-test, partition:1, key:key1560695393135, value:key1560695393135：value9

        根据控制台结果，可以看到一共生产了10条消息，其中根据分区规则，有6条消息保存在partition分区1,4条在分区0，如果数据量大，两个分区的数据是很平均的。（想要多个生产者的就多开几个线程去生产消息）

        4：编码消费者类DConsumer.java：

import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Properties;

import kafka.consumer.Consumer;
import kafka.consumer.ConsumerConfig;
import kafka.consumer.ConsumerIterator;
import kafka.consumer.KafkaStream;
import kafka.javaapi.consumer.ConsumerConnector;
import kafka.message.MessageAndMetadata;

public class DConsumer extends Thread{
	public static void main(String[] args) throws InterruptedException {
		//开两个线程，两个连接，相当于两个消费者
		DConsumer consumer1 = new DConsumer("consumer-1");//消费者consumer-1
		consumer1.start();
		DConsumer consumer2 = new DConsumer("consumer-2");//消费者consumer-2
		consumer2.start();
	}
	private String consumerName = null;
	public DConsumer ( String consumerName) {
		this.consumerName = consumerName;
	} 
    @Override
    public void run() {
    	System.out.println("消费者：" +consumerName + "开始处理消息");
		String topic = "dgh-test";
		Properties props = new Properties();
		props.put("group.id", "group1");
		//props.put("zookeeper.connect", "127.0.0.132:2181,127.0.0.133:2182,127.0.0.134:2183");
		props.put("zookeeper.connect", "127.0.0.1:2181");//连接 zookeeper
		props.put("zookeeper.session.timeout.ms", "400");
		props.put("zookeeper.sync.time.ms", "200");
		props.put("auto.commit.interval.ms", "1000");
		props.put("auto.offset.reset", "smallest"); 
		ConsumerConnector consumer = Consumer.createJavaConsumerConnector(new ConsumerConfig(props));
		Map<String, Integer> topicCountMap = new HashMap<String, Integer>();
		topicCountMap.put(topic, new Integer(1));
		Map<String, List<KafkaStream<byte[], byte[]>>> consumerMap = consumer.createMessageStreams(topicCountMap);
		KafkaStream<byte[], byte[]> stream = consumerMap.get(topic).get(0);
		ConsumerIterator<byte[], byte[]> it = stream.iterator();
		while (it.hasNext()) {
			MessageAndMetadata<byte[], byte[]> mam = it.next();
			System.out.println("consumer: " + consumerName + ", Partition: " + mam.partition() + ", Message: " + new String(mam.message()) + "");
			try {
				sleep(3000);
			} catch (InterruptedException e) { 
				e.printStackTrace();
			}
		}
	}
}

        右键运行：

消费者：consumer-1开始处理消息
消费者：consumer-2开始处理消息
[2019-06-16 22:38:40,317] INFO Starting ZkClient event thread. (org.I0Itec.zkclient.ZkEventThread)
[2019-06-16 22:38:40,317] INFO Starting ZkClient event thread. (org.I0Itec.zkclient.ZkEventThread)
consumer: consumer-1, Partition: 0, Message: key1560695393127：value3
consumer: consumer-2, Partition: 1, Message: key1560695392884：value0
consumer: consumer-2, Partition: 1, Message: key1560695393122：value1
consumer: consumer-1, Partition: 0, Message: key1560695393129：value5
consumer: consumer-1, Partition: 0, Message: key1560695393130：value6
consumer: consumer-2, Partition: 1, Message: key1560695393124：value2
consumer: consumer-2, Partition: 1, Message: key1560695393128：value4
consumer: consumer-1, Partition: 0, Message: key1560695393132：value7
consumer: consumer-2, Partition: 1, Message: key1560695393133：value8
consumer: consumer-2, Partition: 1, Message: key1560695393135：value9

        在消费者连接进行消息消费的时候，如果只有一个消费者，那两个分区的消息都是会把消息发送给这个消费者，如果两个消费者的话，先连接的消费者获取的分区0的消息，后连接的消费者获取分区1的消息。如果三个消费者的话，第三个连的消费者会获取不了消息，所以消费者的数量最好不要超过分区数。