前言：

Flink的DataSoures模块中，定义了DataStream API的数据输入操作，Flink将数据源主要分为内置和第三方数据源，内置数据源包含文件、Socket网络端口以及集合数据类型，主要用于数据测试。而第三方数据源，Flink通过SourceFunction定义了很多的数据源连接器，通过它们进行数据的读写操作。本篇通过Flink的kafka connector实现消费效果。

flink消费kafka:

引入依赖：

		<dependency>
			<groupId>org.springframework.boot</groupId>
			<artifactId>spring-boot-starter-web</artifactId>
		</dependency>
		<dependency>
			<groupId>org.springframework.kafka</groupId>
			<artifactId>spring-kafka</artifactId>
			<version>2.1.0.RELEASE</version>
		</dependency>

		<dependency>
			<groupId>org.apache.flink</groupId>
			<artifactId>flink-connector-kafka_2.11</artifactId>
			<version>1.10.0</version>
		</dependency>

		<dependency>
			<groupId>org.apache.flink</groupId>
			<artifactId>flink-java</artifactId>
			<version>1.10.0</version>
		</dependency>
		<dependency>
			<groupId>org.apache.flink</groupId>
			<artifactId>flink-streaming-java_2.11</artifactId>
			<version>1.10.0</version>
		</dependency>
		<dependency>
			<groupId>org.apache.flink</groupId>
			<artifactId>flink-clients_2.11</artifactId>
			<version>1.10.0</version>
		</dependency>
		<dependency>
			<groupId>joda-time</groupId>
			<artifactId>joda-time</artifactId>
			<version>2.8.1</version>
		</dependency>
		<dependency>
			<groupId>com.alibaba</groupId>
			<artifactId>fastjson</artifactId>
			<version>1.2.73</version>
		</dependency>
		<dependency>
			<groupId>org.projectlombok</groupId>
			<artifactId>lombok</artifactId>
			<optional>true</optional>
		</dependency>
		<dependency>
			<groupId>org.springframework.boot</groupId>
			<artifactId>spring-boot-starter-test</artifactId>
			<scope>test</scope>
		</dependency>

自定义消息序列化类：

import org.apache.flink.api.common.typeinfo.TypeHint;
import org.apache.flink.api.common.typeinfo.TypeInformation;
import org.apache.flink.streaming.connectors.kafka.KafkaDeserializationSchema;
import org.apache.kafka.clients.consumer.ConsumerRecord;

public class MessageDeSerializationSchema implements KafkaDeserializationSchema<ConsumerRecord<String, String>> {

    @Override
    public boolean isEndOfStream(ConsumerRecord<String, String> stringStringConsumerRecord) {
        return false;
    }

    @Override
    public ConsumerRecord<String, String> deserialize(ConsumerRecord<byte[], byte[]> consumerRecord) throws Exception {
        //返回ConsumerRecord
        return new ConsumerRecord<>(
                consumerRecord.topic(),
                consumerRecord.partition(),
                consumerRecord.offset(),
                new String(consumerRecord.key()),
                new String(consumerRecord.value())
        );
    }

    @Override
    public TypeInformation<ConsumerRecord<String, String>> getProducedType() {
        return TypeInformation.of(new TypeHint<ConsumerRecord<String, String>>(){});
    }
}

消费程序：

import org.apache.flink.api.common.functions.FlatMapFunction;
import org.apache.flink.streaming.api.datastream.DataStreamSource;
import org.apache.flink.streaming.api.CheckpointingMode;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.streaming.connectors.kafka.FlinkKafkaConsumer;
import org.apache.flink.util.Collector;
import org.apache.kafka.clients.consumer.ConsumerRecord;

import java.util.Properties;

public class FlinkConsumer {

    public static void main(String[] args) throws Exception{
        // 构建环境
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        //指定checkpoint的触发间隔
        env.enableCheckpointing(5000);
        // 默认的CheckpointingMode就是EXACTLY_ONCE，也可以指定为AT_LEAST_ONCE
		env.getCheckpointConfig().setCheckpointingMode(CheckpointingMode.EXACTLY_ONCE);
		
        Properties properties = new Properties();

        properties.setProperty("bootstrap.servers", "192.168.9.239:9092");
        //设置消费者组
        properties.setProperty("group.id", "flink_consumer");
        //消费的三种方式，默认是latest
        //earliest：各分区下有已提交的offset时，从提交的offset开始消费；无提交的offset时，从头开始消费
        //latest：各分区下有已提交的offset时，从提交的offset开始消费；无提交的offset时，消费新产生的该分区下的数据
        //none： topic各分区都存在已提交的offset时，从offset后开始消费；只要有一个分区不存在已提交的offset，则抛出异常
        properties.setProperty("auto.offset.reset", "earliest");

        FlinkKafkaConsumer kafkaConsumer = new FlinkKafkaConsumer<>("flink_kafka", new MessageDeSerializationSchema(), properties);
        
        DataStreamSource<ConsumerRecord<String, String>> stream = env
                .addSource(kafkaConsumer).setParallelism(2);

        //消费多个topic
//        List<String> topics = new ArrayList<>();
//        topics.add("topic1");
//        topics.add("topic2");
//        DataStream<String> stream = env
//                .addSource(new FlinkKafkaConsumer<>(topics, new SimpleStringSchema(), properties));


        stream.flatMap(new FlatMapFunction<ConsumerRecord<String, String>, Object>() {
            @Override
            public void flatMap(ConsumerRecord<String, String> record, Collector<Object> collector) throws Exception {
                collector.collect("offset:"+record.offset());
                collector.collect("partition:"+record.partition());
                collector.collect("topic:"+record.topic());
                collector.collect("key:"+record.key());
                collector.collect("value:"+record.value());
            }
        }).print();

        env.execute("consumer start");
    }
}

消息生产程序：

import com.alibaba.fastjson.JSON;
import com.tt.flinkdemo.domain.MessageEntity;
import lombok.extern.slf4j.Slf4j;
import org.joda.time.DateTime;
import org.springframework.stereotype.Component;
import org.apache.kafka.clients.producer.KafkaProducer;
import org.apache.kafka.clients.producer.ProducerRecord;
import org.apache.kafka.clients.producer.RecordMetadata;

import java.util.Properties;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.Future;


@Slf4j
@Component
public class MessageProducter {

    private static Properties getProps(){
        Properties props =  new Properties();
        props.put("bootstrap.servers", "192.168.9.239:9092");
        props.put("acks", "all"); // 发送所有ISR
        props.put("retries", Integer.MAX_VALUE); // 重试次数
        props.put("batch.size", 16384); // 批量发送大小
        props.put("buffer.memory", 102400); // 缓存大小，根据本机内存大小配置
        props.put("linger.ms", 1000); // 发送频率，满足任务一个条件发送
        props.put("client.id", "producer-syn-1"); // 发送端id,便于统计
        props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");
        props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");
        return props;
    }


    public static void main(String[] args) throws Exception {
        KafkaProducer<String, String> producer = new KafkaProducer<>(getProps());
        for (int i=20;i<=30;i++){
            MessageEntity message = new MessageEntity();
            message.setMessage("第"+i+"条：message");
            message.setTotalDate(DateTime.now().toString("yyyy-MM-dd HH:mm:ss"));
            message.setMessageId(System.currentTimeMillis());
            log.info(JSON.toJSONString(message));
            ProducerRecord<String, String> record = new ProducerRecord<>("flink_kafka",
                    System.currentTimeMillis()+"", JSON.toJSONString(message));
            Future<RecordMetadata> metadataFuture = producer.send(record);
            RecordMetadata recordMetadata;
            try {
                recordMetadata = metadataFuture.get();
                log.info("发送成功！");
                log.info("topic:"+recordMetadata.topic());
                log.info("partition:"+recordMetadata.partition());
                log.info("offset:"+recordMetadata.offset());
            } catch (InterruptedException|ExecutionException e) {
                System.out.println("发送失败！");
                e.printStackTrace();
            }
            Thread.sleep(100);
        }
    }
}

测试效果：