Flink最后一站___Flink数据写入Kafka+从Kafka存入Mysql

Maynor

发布于 2021-12-07 10:29:28

1.2K0

发布于 2021-12-07 10:29:28

文章被收录于专栏：最新最全的大数据技术体系最新最全的大数据技术体系

前言

大家好,我是ChinaManor,直译过来就是中国码农的意思,我希望自己能成为国家复兴道路的铺路人,大数据领域的耕耘者,平凡但不甘于平庸的人。

今天为大家带来Flink的一个综合应用案例:Flink数据写入Kafka+从Kafka存入Mysql 第一部分:写数据到kafka中

 public static void writeToKafka() throws Exception{
        Properties props = new Properties();
        props.put("bootstrap.servers", BROKER_LIST);
        props.put("key.serializer", CONST_SERIALIZER);
        props.put("value.serializer", CONST_SERIALIZER);

        KafkaProducer<String, String> producer = new KafkaProducer<>(props);

        //构建User对象，在name为data后边加个随机数
        int randomInt = RandomUtils.nextInt(1, 100000);
        User user = new User();
        user.setName("data" + randomInt);
        user.setId(randomInt);
        //转换成JSON
        String userJson = JSON.toJSONString(user);

        //包装成kafka发送的记录
        ProducerRecord<String, String> record = new ProducerRecord<String, String>(TOPIC_USER, partition,
                null, userJson);
        //发送到缓存
        producer.send(record);
        System.out.println("向kafka发送数据:" + userJson);
        //立即发送
        producer.flush();

    }

重点:

//发送到缓存
        producer.send(record);

为了增强代码的Robust,我们将常量单独拎出来:

   //本地的kafka机器列表
    public static final String BROKER_LIST = "192.168.88.161:9092";
    //kafka的topic
    public static final String TOPIC_USER = "USER";
    //kafka的partition分区
    public static final Integer partition = 0;

    //序列化的方式
    public static final String CONST_SERIALIZER = "org.apache.kafka.common.serialization.StringSerializer";
    //反序列化
    public static final String CONST_DESERIALIZER = "org.apache.kafka.common.serialization.StringDeserializer";

main方法如下:

public static void main(String[] args) {
        while(true) {
            try {
                //每三秒写一条数据
                TimeUnit.SECONDS.sleep(3);
                writeToKafka();
            } catch (Exception e) {
                e.printStackTrace();
            }

        }
    }

第二部分:从kafka获取数据

KafkaRickSourceFunction.java

import com.hy.flinktest.entity.User;
import lombok.extern.slf4j.Slf4j;
import org.apache.flink.api.java.tuple.Tuple2;
import org.apache.flink.streaming.api.functions.source.RichSourceFunction;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.apache.kafka.clients.consumer.KafkaConsumer;
import org.apache.kafka.common.TopicPartition;
import org.apache.zookeeper.WatchedEvent;
import org.apache.zookeeper.Watcher;
import org.apache.zookeeper.ZooKeeper;

import java.io.IOException;
import java.time.Duration;
import java.util.Collections;
import java.util.List;
import java.util.Properties;


@Slf4j
public class KafkaRickSourceFunction extends RichSourceFunction<String>{
    //kafka
    private static Properties prop = new Properties();
    private boolean running = true;

	//作静态化处理,增强robust
    private static Integer partition = WritedatatoKafka.partition;
    static {
        prop.put("bootstrap.servers",WritedatatoKafka.BROKER_LIST);
        prop.put("zookeeper.connect","192.168.88.161:2181");
        prop.put("group.id",WritedatatoKafka.TOPIC_USER);
        prop.put("key.deserializer",WritedatatoKafka.CONST_DESERIALIZER);
        prop.put("value.deserializer",WritedatatoKafka.CONST_DESERIALIZER);
        prop.put("auto.offset.reset","latest");
        prop.put("max.poll.records", "500");
        prop.put("auto.commit.interval.ms", "1000");
    }

    @Override
    public void run(SourceContext sourceContext) throws Exception {
        //创建一个消费者客户端实例
        KafkaConsumer<String,String> kafkaConsumer = new KafkaConsumer<String, String>(prop);
        //只消费TOPIC_USER 分区
        TopicPartition topicPartition = new TopicPartition(WritedatatoKafka.TOPIC_USER,partition);
        long offset =0; //这个初始值应该从zk或其他地方获取
        offset = placeOffsetToBestPosition(kafkaConsumer, offset, topicPartition);


        while (running){
            ConsumerRecords<String, String> records = kafkaConsumer.poll(1000);
            if(records.isEmpty()){
                continue;
            }
            for (ConsumerRecord<String, String> record : records) {
                //record.offset();
                //record.key()
                String value = record.value();
                sourceContext.collect(value);
            }
        }

    }

然后返回最合适的offset

    /**
     * 将offset定位到最合适的位置，并返回最合适的offset。
     * @param kafkaConsumer consumer
     * @param offset offset
     * @param topicPartition partition
     * @return the best offset
     */
    private long placeOffsetToBestPosition(
            KafkaConsumer<String, String> kafkaConsumer,
            long offset, TopicPartition topicPartition) {
        List<TopicPartition> partitions = Collections.singletonList(topicPartition);
        kafkaConsumer.assign(partitions);
        long bestOffset = offset;
        if (offset == 0) {
            log.info("由于offset为0，重新定位offset到kafka起始位置.");
            kafkaConsumer.seekToBeginning(partitions);

        } else if (offset > 0) {

            kafkaConsumer.seekToBeginning(partitions);
            long startPosition = kafkaConsumer.position(topicPartition);
            kafkaConsumer.seekToEnd(partitions);
            long endPosition = kafkaConsumer.position(topicPartition);

            if (offset < startPosition) {
                log.info("由于当前offset({})比kafka的最小offset({})还要小，则定位到kafka的最小offset({})处。",
                        offset, startPosition, startPosition);
                kafkaConsumer.seekToBeginning(partitions);
                bestOffset = startPosition;
            } else if (offset > endPosition) {
                log.info("由于当前offset({})比kafka的最大offset({})还要大，则定位到kafka的最大offset({})处。",
                        offset, endPosition, endPosition);
                kafkaConsumer.seekToEnd(partitions);
                bestOffset = endPosition;
            } else {
                kafkaConsumer.seek(topicPartition, offset);
            }
        }
        return bestOffset;
    }

    @Override
    public void cancel() {
        running = false;
    }

}

第三部分主类:从kafka读取数据写入mysql

    //1.构建流执行环境 并添加数据源

		StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        DataStreamSource<String> dataStreamSource = env.addSource(new KafkaRickSourceFunction());

    //2.从kafka里读取数据，转换成User对象

 DataStream<User> dataStream = dataStreamSource.map(lines -> JSONObject.parseObject(lines, User.class));

//3.收集5秒钟的总数

dataStream.timeWindowAll(Time.seconds(5L)).
        apply(new AllWindowFunction<User, List<User>, TimeWindow>() {

            @Override
            public void apply(TimeWindow timeWindow, Iterable<User> iterable, Collector<List<User>> out) throws Exception {
                List<User> users = Lists.newArrayList(iterable);

                if(users.size() > 0) {
                    System.out.println("5秒内总共收到的条数：" + users.size());
                    out.collect(users);
                }

            }
        })
        //sink 到数据库
                .addSink(new MysqlRichSinkFunction());
        //打印到控制台
        //.print();

第四部分: 写入到目标数据库sink MysqlRichSinkFunction.java

@Slf4j
public class MysqlRichSinkFunction extends RichSinkFunction<List<User>> {

    private Connection connection = null;
    private PreparedStatement ps = null;

    @Override
    public void open(Configuration parameters) throws Exception {
       // super.open(parameters);
        log.info("获取数据库连接");
        connection = DbUtil.getConnection();
        String sql = "insert into user1(id,name) values (?,?)";
        ps = connection.prepareStatement(sql);
    }

    public void invoke(List<User> users, Context ctx) throws Exception {
        //获取ReadMysqlResoure发送过来的结果
        for(User user : users) {
            ps.setLong(1, user.getId());
            ps.setString(2, user.getName());
            ps.addBatch();
        }
        //一次性写入
        int[] count = ps.executeBatch();
        log.info("成功写入Mysql数量：" + count.length);

    }


    @Override
    public void close() throws Exception {
        //关闭并释放资源
        if(connection != null) {
            connection.close();
        }

        if(ps != null) {
            ps.close();
        }
    }

}

本文参与?腾讯云自媒体同步曝光计划，分享自作者个人站点/博客。

原始发表：2021-07-13 ，如有侵权请联系 cloudcommunity@tencent.com 删除

flink