kafka的简单使用

为什么80%的码农都做不了架构师？>>>   

在eclipse中新建kafka-demo的maven项目，pom.xml依赖如下

<dependencies>
    <dependency>
      <groupId>junit</groupId>
      <artifactId>junit</artifactId>
      <version>4.11</version>
      <scope>test</scope>
    </dependency>
    
    <dependency>
		<groupId>org.apache.kafka</groupId>
		<artifactId>kafka_2.9.1</artifactId>
		<version>0.8.2.1</version>
	</dependency>
	
	<!-- <dependency>
		<groupId>org.apache.kafka</groupId>
		<artifactId>kafka-clients</artifactId>
		<version>0.8.2.1</version>
	</dependency> -->
    
    
    
  </dependencies>

ProducerDemo.java

package com.leech.kafka.demo;

import java.util.Date;
import java.util.Properties;

import kafka.javaapi.producer.Producer;
import kafka.producer.KeyedMessage;
import kafka.producer.ProducerConfig;


public class ProducerDemo {
	 public static void main(String[] args) {
	        //Random rnd = new Random();
	        int events=100;
	 
	        // 设置配置属性
	        Properties props = new Properties();
	        //props.put("metadata.broker.list","172.168.63.221:9092,172.168.63.233:9092,172.168.63.234:9092");
	        props.put("metadata.broker.list","192.168.1.82:9092");
	        props.put("serializer.class", "kafka.serializer.StringEncoder");
	        // key.serializer.class默认为serializer.class
	        props.put("key.serializer.class", "kafka.serializer.StringEncoder");
	        // 可选配置，如果不配置，则使用默认的partitioner
	        props.put("partitioner.class", "com.leech.kafka.demo.PartitionerDemo");
	        // 触发acknowledgement机制，否则是fire and forget，可能会引起数据丢失
	        // 值为0,1,-1,可以参考
	        // http://kafka.apache.org/08/configuration.html
	        props.put("request.required.acks", "1");
	        ProducerConfig config = new ProducerConfig(props);
	 
	        // 创建producer
	        Producer<String, String> producer = new Producer<String, String>(config);
	        // 产生并发送消息
	        long start=System.currentTimeMillis();
	        for (long i = 0; i < events; i++) {
	            long runtime = new Date().getTime();
	            String ip = "192.168.2." + i;//rnd.nextInt(255);
	            String msg = runtime + ",www.example.com," + ip;
	            //如果topic不存在，则会自动创建，默认replication-factor为1，partitions为0
	            KeyedMessage<String, String> data = new KeyedMessage<String, String>("page_visits", ip, msg);
	            producer.send(data);
	        }
	        System.out.println("耗时:" + (System.currentTimeMillis() - start));
	        // 关闭producer
	        producer.close();
	    }
}

ConsumerDemo.java

package com.leech.kafka.demo;

import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Properties;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
 
import kafka.consumer.Consumer;
import kafka.consumer.ConsumerConfig;
import kafka.consumer.KafkaStream;
import kafka.javaapi.consumer.ConsumerConnector;
 
/**
 * 详细可以参考：https://cwiki.apache.org/confluence/display/KAFKA/Consumer+Group+Example
 * 
 * @author Fung
 *
 */
public class ConsumerDemo {
    private final ConsumerConnector consumer;
    private final String topic;
    private ExecutorService executor;
 
    public ConsumerDemo(String a_zookeeper, String a_groupId, String a_topic) {
        consumer = Consumer.createJavaConsumerConnector(createConsumerConfig(a_zookeeper,a_groupId));
        this.topic = a_topic;
    }
 
    public void shutdown() {
        if (consumer != null)
            consumer.shutdown();
        if (executor != null)
            executor.shutdown();
    }
 
    public void run(int numThreads) {
        Map<String, Integer> topicCountMap = new HashMap<String, Integer>();
        topicCountMap.put(topic, new Integer(numThreads));
        Map<String, List<KafkaStream<byte[], byte[]>>> consumerMap = consumer
                .createMessageStreams(topicCountMap);
        List<KafkaStream<byte[], byte[]>> streams = consumerMap.get(topic);
 
        // now launch all the threads
        executor = Executors.newFixedThreadPool(numThreads);
 
        // now create an object to consume the messages
        //
        int threadNumber = 0;
        for (final KafkaStream stream : streams) {
            executor.submit(new ConsumerMsgTask(stream, threadNumber));
            threadNumber++;
        }
    }
 
    private static ConsumerConfig createConsumerConfig(String a_zookeeper,
            String a_groupId) {
        Properties props = new Properties();
        props.put("zookeeper.connect", a_zookeeper);
        props.put("group.id", a_groupId);
        props.put("zookeeper.session.timeout.ms", "400");
        props.put("zookeeper.sync.time.ms", "200");
        props.put("auto.commit.interval.ms", "1000");
 
        return new ConsumerConfig(props);
    }
 
    public static void main(String[] arg) {
        String[] args = { "192.168.1.82:2181", "group-1", "page_visits", "12" };
        String zooKeeper = args[0];
        String groupId = args[1];
        String topic = args[2];
        int threads = Integer.parseInt(args[3]);
 
        ConsumerDemo demo = new ConsumerDemo(zooKeeper, groupId, topic);
        demo.run(threads);
 
        try {
            Thread.sleep(10000);
        } catch (InterruptedException ie) {
 
        }
        demo.shutdown();
    }
}

ConsumerMsgTask.java

package com.leech.kafka.demo;

import kafka.consumer.ConsumerIterator;
import kafka.consumer.KafkaStream;

public class ConsumerMsgTask implements Runnable {

	 private KafkaStream m_stream;
	    private int m_threadNumber;
	 
	    public ConsumerMsgTask(KafkaStream stream, int threadNumber) {
	        m_threadNumber = threadNumber;
	        m_stream = stream;
	    }
	 
	    @Override
	    public void run() {
	        ConsumerIterator<byte[], byte[]> it = m_stream.iterator();
	        while (it.hasNext())
	            System.out.println("Thread " + m_threadNumber + ": "
	                    + new String(it.next().message()));
	        System.out.println("Shutting down Thread: " + m_threadNumber);
	    }
}

PartitionerDemo.java

package com.leech.kafka.demo;

import kafka.producer.Partitioner;
import kafka.utils.VerifiableProperties;

public class PartitionerDemo implements Partitioner {

	public PartitionerDemo(VerifiableProperties props) {
		 
    }
 
    @Override
    public int partition(Object obj, int numPartitions) {
        int partition = 0;
        if (obj instanceof String) {
            String key=(String)obj;
            int offset = key.lastIndexOf('.');
            if (offset > 0) {
                partition = Integer.parseInt(key.substring(offset + 1)) % numPartitions;
            }
        }else{
            partition = obj.toString().length() % numPartitions;
        }
         
        return partition;
    }

}

转载于:https://my.oschina.net/chaun/blog/408511