Kafka API实战

1 环境准备

1）在eclipse中创建一个java工程

2）在工程的根目录创建一个lib文件夹

3）解压kafka安装包，将安装包libs目录下的jar包拷贝到工程的lib目录下，并build path。

4）启动zk和kafka集群，在kafka集群中打开一个消费者

[atguigu@hadoop102 kafka]$ bin/kafka-console-consumer.sh --zookeeper hadoop102:2181 --topic first

2 Kafka生产者Java API

2.1 创建生产者（过时的API）

package com.atguigu.kafka; import java.util.Properties; import kafka.javaapi.producer.Producer; import kafka.producer.KeyedMessage; import kafka.producer.ProducerConfig;   public class OldProducer {          @SuppressWarnings("deprecation")        public static void main(String[] args) {                             Properties properties = new Properties();               properties.put("metadata.broker.list", "hadoop102:9092");               properties.put("request.required.acks", "1");               properties.put("serializer.class", "kafka.serializer.StringEncoder");                             Producer<Integer, String> producer = new Producer<Integer,String>(new ProducerConfig(properties));                             KeyedMessage<Integer, String> message = new KeyedMessage<Integer, String>("first", "hello world");               producer.send(message );        } }

2.2 创建生产者（新API）

package com.atguigu.kafka; import java.util.Properties; import org.apache.kafka.clients.producer.KafkaProducer; import org.apache.kafka.clients.producer.Producer; import org.apache.kafka.clients.producer.ProducerRecord;   public class NewProducer {          public static void main(String[] args) {                             Properties props = new Properties();               // Kafka服务端的主机名和端口号               props.put("bootstrap.servers", "hadoop103:9092");               // 等待所有副本节点的应答               props.put("acks", "all");               // 消息发送最大尝试次数               props.put("retries", 0);               // 一批消息处理大小               props.put("batch.size", 16384);               // 请求延时               props.put("linger.ms", 1);               // 发送缓存区内存大小               props.put("buffer.memory", 33554432);               // key序列化               props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");               // value序列化               props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");                 KafkaProducer<String, String> producer = new KafkaProducer<>(props);               for (int i = 0; i < 50; i++) {                      producer.send(new ProducerRecord<String, String>("first", Integer.toString(i), "hello world-" + i));               }                 producer.close();        } }

 

2.3 创建生产者带回调函数（新API）

package com.atguigu.kafka; import java.util.Properties; import org.apache.kafka.clients.producer.Callback; import org.apache.kafka.clients.producer.KafkaProducer; import org.apache.kafka.clients.producer.ProducerRecord; import org.apache.kafka.clients.producer.RecordMetadata;   public class CallBackProducer {          public static void main(String[] args) {   Properties props = new Properties();               // Kafka服务端的主机名和端口号               props.put("bootstrap.servers", "hadoop103:9092");               // 等待所有副本节点的应答               props.put("acks", "all");               // 消息发送最大尝试次数               props.put("retries", 0);               // 一批消息处理大小               props.put("batch.size", 16384);               // 增加服务端请求延时               props.put("linger.ms", 1); // 发送缓存区内存大小               props.put("buffer.memory", 33554432);               // key序列化               props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");               // value序列化               props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");                 KafkaProducer<String, String> kafkaProducer = new KafkaProducer<>(props);                 for (int i = 0; i < 50; i++) {                        kafkaProducer.send(new ProducerRecord<String, String>("first", "hello" + i), new Callback() {                               @Override                             public void onCompletion(RecordMetadata metadata, Exception exception) {                                      if (metadata != null) {                                             System.out.println(metadata.partition() + "---" + metadata.offset());                                    }                             }                      });               }                 kafkaProducer.close();        } }

 

2.4 自定义分区生产者

0）需求：将所有数据存储到topic的第0号分区上

1）定义一个类实现Partitioner接口，重写里面的方法（过时API）

package com.atguigu.kafka; import java.util.Map; import kafka.producer.Partitioner;   public class CustomPartitioner implements Partitioner {          public CustomPartitioner() {               super();        }          @Override        public int partition(Object key, int numPartitions) {               // 控制分区               return 0;        } }

2）自定义分区（新API）

package com.atguigu.kafka; import java.util.Map; import org.apache.kafka.clients.producer.Partitioner; import org.apache.kafka.common.Cluster;   public class CustomPartitioner implements Partitioner {          @Override        public void configure(Map<String, ?> configs) {                      }          @Override        public int partition(String topic, Object key, byte[] keyBytes, Object value, byte[] valueBytes, Cluster cluster) {         // 控制分区               return 0;        }          @Override        public void close() {                      } }

3）在代码中调用

package com.atguigu.kafka; import java.util.Properties; import org.apache.kafka.clients.producer.KafkaProducer; import org.apache.kafka.clients.producer.Producer; import org.apache.kafka.clients.producer.ProducerRecord;   public class PartitionerProducer {          public static void main(String[] args) {                             Properties props = new Properties();               // Kafka服务端的主机名和端口号               props.put("bootstrap.servers", "hadoop103:9092");               // 等待所有副本节点的应答               props.put("acks", "all");               // 消息发送最大尝试次数               props.put("retries", 0);               // 一批消息处理大小               props.put("batch.size", 16384);               // 增加服务端请求延时               props.put("linger.ms", 1);               // 发送缓存区内存大小               props.put("buffer.memory", 33554432);               // key序列化               props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");               // value序列化               props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");               // 自定义分区               props.put("partitioner.class", "com.atguigu.kafka.CustomPartitioner");                 Producer<String, String> producer = new KafkaProducer<>(props);               producer.send(new ProducerRecord<String, String>("first", "1", "atguigu"));                 producer.close();        } }

4）测试

       （1）在hadoop102上监控/opt/module/kafka/logs/目录下first主题3个分区的log日志动态变化情况

[atguigu@hadoop102 first-0]$ tail -f 00000000000000000000.log

[atguigu@hadoop102 first-1]$ tail -f 00000000000000000000.log

[atguigu@hadoop102 first-2]$ tail -f 00000000000000000000.log

       （2）发现数据都存储到指定的分区了。

3 Kafka消费者Java API

0）在控制台创建发送者

[atguigu@hadoop104 kafka]$ bin/kafka-console-producer.sh --broker-list hadoop102:9092 --topic first

>hello world

1）创建消费者（过时API）

package com.atguigu.kafka.consume; 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;   public class CustomConsumer {          @SuppressWarnings("deprecation")        public static void main(String[] args) {               Properties properties = new Properties();                             properties.put("zookeeper.connect", "hadoop102:2181");               properties.put("group.id", "g1");               properties.put("zookeeper.session.timeout.ms", "500");               properties.put("zookeeper.sync.time.ms", "250");               properties.put("auto.commit.interval.ms", "1000");                             // 创建消费者连接器               ConsumerConnector consumer = Consumer.createJavaConsumerConnector(new ConsumerConfig(properties));                             HashMap<String, Integer> topicCount = new HashMap<>();               topicCount.put("first", 1);                             Map<String, List<KafkaStream<byte[], byte[]>>> consumerMap = consumer.createMessageStreams(topicCount);                             KafkaStream<byte[], byte[]> stream = consumerMap.get("first").get(0);                             ConsumerIterator<byte[], byte[]> it = stream.iterator();                             while (it.hasNext()) {                      System.out.println(new String(it.next().message()));               }        } }

2）官方提供案例（自动维护消费情况）（新API）

package com.atguigu.kafka.consume; import java.util.Arrays; import java.util.Properties; import org.apache.kafka.clients.consumer.ConsumerRecord; import org.apache.kafka.clients.consumer.ConsumerRecords; import org.apache.kafka.clients.consumer.KafkaConsumer;   public class CustomNewConsumer {          public static void main(String[] args) {                 Properties props = new Properties();               // 定义kakfa 服务的地址，不需要将所有broker指定上               props.put("bootstrap.servers", "hadoop102:9092");               // 制定consumer group               props.put("group.id", "test");               // 是否自动确认offset               props.put("enable.auto.commit", "true");               // 自动确认offset的时间间隔               props.put("auto.commit.interval.ms", "1000");               // key的序列化类               props.put("key.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");               // value的序列化类               props.put("value.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");               // 定义consumer               KafkaConsumer<String, String> consumer = new KafkaConsumer<>(props);                             // 消费者订阅的topic, 可同时订阅多个               consumer.subscribe(Arrays.asList("first", "second","third"));                 while (true) {                      // 读取数据，读取超时时间为100ms                      ConsumerRecords<String, String> records = consumer.poll(100);                                           for (ConsumerRecord<String, String> record : records)                             System.out.printf("offset = %d, key = %s, value = %s%n", record.offset(), record.key(), record.value());               }        } }