Flink清洗Kafka数据存入MySQL测试

版本信息：

Flink Version:1.6.2
Kafka Version:0.9.0.0
MySQL Version:5.6.21

Kafka 消息样例及格式：[IP TIME URL STATU_CODE REFERER]

1.74.103.143	2018-12-20 18:12:00	 "GET /class/130.html HTTP/1.1" 	404	https://search.yahoo.com/search?p=Flink实战

Pom.xml

2.11.8
1.6.2

 
      org.apache.flink
      flink-java
      ${flink.version}
    

    
      org.apache.flink
      flink-streaming-java_2.11
      ${flink.version}
    
    
      org.apache.flink
      flink-clients_2.11
      ${flink.version}
    

    
    
      org.apache.flink
      flink-connector-kafka-0.9_2.11
      ${flink.version}
    

    
      mysql
      mysql-connector-java
      5.1.39
    

sConf

package com.soul.conf;

/**
 * @author soulChun
 * @create 2018-12-20-15:11
 */
public class sConf {
    public static final String USERNAME = "root";
    public static final String PASSWORD = "root";
    public static final String DRIVERNAME = "com.mysql.jdbc.Driver";
    public static final String URL = "jdbc:mysql://localhost:3306/soul";
}

MySQLSlink

package com.soul.kafka;

import com.soul.conf.sConf;
import org.apache.flink.api.java.tuple.Tuple5;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.streaming.api.functions.sink.RichSinkFunction;

import java.sql.Connection;
import java.sql.DriverManager;
import java.sql.PreparedStatement;

/**
 * @author soulChun
 * @create 2018-12-20-15:09
 */
public class MySQLSink extends RichSinkFunction> {
    private static final long serialVersionUID = 1L;
    private Connection connection;
    private PreparedStatement preparedStatement;

    public void invoke(Tuple5 value) {

        try {
            if (connection == null) {
                Class.forName(sConf.DRIVERNAME);
                connection = DriverManager.getConnection(sConf.URL, sConf.USERNAME, sConf.PASSWORD);
            }
            String sql = "insert into log_info (ip,time,courseid,status_code,referer) values (?,?,?,?,?)";
            preparedStatement = connection.prepareStatement(sql);
            preparedStatement.setString(1, value.f0);
            preparedStatement.setString(2, value.f1);
            preparedStatement.setString(3, value.f2);
            preparedStatement.setString(4, value.f3);
            preparedStatement.setString(5, value.f4);
            System.out.println("Start insert");
            preparedStatement.executeUpdate();
        } catch (Exception e) {
            e.printStackTrace();
        }
    }

    public void open(Configuration parms) throws Exception {
        Class.forName(sConf.DRIVERNAME);
        connection = DriverManager.getConnection(sConf.URL, sConf.USERNAME, sConf.PASSWORD);
    }

    public void close() throws Exception {

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

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

    }


}

数据清洗日期工具类

package com.soul.utils;

import org.apache.commons.lang3.time.FastDateFormat;

import java.util.Date;

/**
 * @author soulChun
 * @create 2018-12-19-18:44
 */
public class DateUtils {
    private static FastDateFormat SOURCE_FORMAT = FastDateFormat.getInstance("yyyy-MM-dd HH:mm:ss");
    private static FastDateFormat TARGET_FORMAT = FastDateFormat.getInstance("yyyyMMddHHmmss");

    public static Long  getTime(String  time) throws Exception{
        return SOURCE_FORMAT.parse(time).getTime();
    }

    public static String parseMinute(String time) throws  Exception{
        return TARGET_FORMAT.format(new Date(getTime(time)));
    }


    public static void main(String[] args) throws Exception{
        String time = "2018-12-19 18:55:00";

        System.out.println(parseMinute(time));
    }
}

MySQL建表

create table log_info(
ID INT NOT NULL AUTO_INCREMENT,
IP VARCHAR(50),
TIME VARCHAR(50),
CourseID VARCHAR(10),
Status_Code VARCHAR(10),
Referer VARCHAR(100),
PRIMARY KEY ( ID )
)ENGINE=InnoDB DEFAULT CHARSET=utf8;

主程序：主要是将time的格式转成yyyyMMddHHmmss,还有取URL中的课程ID，将不是/class开头的过滤掉。

package com.soul.kafka;

import com.soul.utils.DateUtils;
import org.apache.flink.api.common.functions.FilterFunction;
import org.apache.flink.api.common.functions.MapFunction;
import org.apache.flink.api.common.serialization.SimpleStringSchema;
import org.apache.flink.api.java.tuple.Tuple5;
import org.apache.flink.streaming.api.TimeCharacteristic;
import org.apache.flink.streaming.api.datastream.DataStream;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.streaming.connectors.kafka.FlinkKafkaConsumer09;

import java.util.Properties;

/**
 * @author soulChun
 * @create 2018-12-19-17:23
 */
public class FlinkCleanKafka {
    public static void main(String[] args) throws Exception {
        final StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.enableCheckpointing(5000);
        Properties properties = new Properties();
        properties.setProperty("bootstrap.servers", "localhost:9092");//kafka的节点的IP或者hostName，多个使用逗号分隔
        properties.setProperty("zookeeper.connect", "localhost:2181");//zookeeper的节点的IP或者hostName，多个使用逗号进行分隔
        properties.setProperty("group.id", "test-consumer-group");//flink consumer flink的消费者的group.id

        FlinkKafkaConsumer09 myConsumer = new FlinkKafkaConsumer09("imooc_topic", new SimpleStringSchema(), properties);

        DataStream stream = env.addSource(myConsumer);
//        stream.print().setParallelism(2);

        DataStream CleanData = stream.map(new MapFunction>() {
            @Override
            public Tuple5 map(String value) throws Exception {
                String[] data = value.split("\\\t");
                String CourseID = null;
                String url = data[2].split("\\ ")[2];
                if (url.startsWith("/class")) {
                    String CourseHTML = url.split("\\/")[2];
                    CourseID = CourseHTML.substring(0, CourseHTML.lastIndexOf("."));
//                    System.out.println(CourseID);
                }

                return Tuple5.of(data[0], DateUtils.parseMinute(data[1]), CourseID, data[3], data[4]);
            }
        }).filter(new FilterFunction>() {
            @Override
            public boolean filter(Tuple5 value) throws Exception {
                return value.f2 != null;
            }
        });


        CleanData.addSink(new MySQLSink());

        env.execute("Flink kafka");
    }
}

启动主程序，查看表数据

mysql> select count(*) from log_info;
+----------+
| count(*) |
+----------+
|    15137 |
+----------+

Kafka过来的消息是我模拟的，一分钟产生100条。以上只是测试代码，对数据的准确性、程序性能没有做考虑，大家可以自己完善。还有WaterMark之类还在测试，测试完会做分享。