Flink+Kafka整合的实例

Flink+Kafka整合实例

1.使用工具Intellig IDEA新建一个maven项目，为项目命名为kafka01。

2.我的pom.xml文件配置如下。

1<?xml version="1.0" encoding="UTF-8"?>
2<project xmlns="http://maven.apache.org/POM/4.0.0"
3         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
4         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
5    <modelVersion>4.0.0</modelVersion>
6
7    <groupId>com.hrb.lhr</groupId>
8    <artifactId>kafka01</artifactId>
9    <version>1.0-SNAPSHOT</version>
10
11    <properties>
12        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
13        <flink.version>1.1.4</flink.version>
14        <slf4j.version>1.7.7</slf4j.version>
15        <log4j.version>1.2.17</log4j.version>
16    </properties>
17
18    <dependencies>
19        <dependency>
20            <groupId>org.apache.flink</groupId>
21            <artifactId>flink-java</artifactId>
22            <version>${flink.version}</version>
23        </dependency>
24        <dependency>
25            <groupId>org.apache.flink</groupId>
26            <artifactId>flink-streaming-java_2.11</artifactId>
27            <version>${flink.version}</version>
28        </dependency>
29        <dependency>
30            <groupId>org.apache.flink</groupId>
31            <artifactId>flink-clients_2.11</artifactId>
32            <version>${flink.version}</version>
33        </dependency>
34        <!-- explicitly add a standard loggin framework, as Flink does not (in the future) have
35           a hard dependency on one specific framework by default -->
36        <dependency>
37            <groupId>org.slf4j</groupId>
38            <artifactId>slf4j-log4j12</artifactId>
39            <version>${slf4j.version}</version>
40        </dependency>
41        <dependency>
42            <groupId>log4j</groupId>
43            <artifactId>log4j</artifactId>
44            <version>${log4j.version}</version>
45        </dependency>
46        <dependency>
47            <groupId>org.apache.flink</groupId>
48            <artifactId>flink-connector-kafka-0.9_2.11</artifactId>
49            <version>${flink.version}</version>
50        </dependency>
51    </dependencies>
52
53</project>

3.在项目的目录/src/main/java在创建两个Java类，分别命名为KafkaDemo和CustomWatermarkEmitter，代码如下所示。

1import java.util.Properties;
2import org.apache.flink.streaming.api.TimeCharacteristic;
3import org.apache.flink.streaming.api.datastream.DataStream;
4import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
5import org.apache.flink.streaming.connectors.kafka.FlinkKafkaConsumer09;
6import org.apache.flink.streaming.util.serialization.SimpleStringSchema;
7
8
9public class KafkaDeme {
10
11        public static void main(String[] args) throws Exception {
12
13                // set up the streaming execution environment
14                final StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
15                //默认情况下，检查点被禁用。要启用检查点，请在StreamExecutionEnvironment上调用enableCheckpointing(n)方法，
16                // 其中n是以毫秒为单位的检查点间隔。每隔5000 ms进行启动一个检查点,则下一个检查点将在上一个检查点完成后5秒钟内启动
17                env.enableCheckpointing(5000);
18                env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);
19                Properties properties = new Properties();
20                properties.setProperty("bootstrap.servers", "10.192.12.106:9092");//kafka的节点的IP或者hostName，多个使用逗号分隔
21                properties.setProperty("zookeeper.connect", "10.192.12.106:2181");//zookeeper的节点的IP或者hostName，多个使用逗号进行分隔
22                properties.setProperty("group.id", "test-consumer-group");//flink consumer flink的消费者的group.id
23                FlinkKafkaConsumer09<String> myConsumer = new FlinkKafkaConsumer09<String>("test0", new SimpleStringSchema(),
24                        properties);//test0是kafka中开启的topic
25                myConsumer.assignTimestampsAndWatermarks(new CustomWatermarkEmitter());
26                DataStream<String> keyedStream = env.addSource(myConsumer);//将kafka生产者发来的数据进行处理，本例子我进任何处理
27                keyedStream.print();//直接将从生产者接收到的数据在控制台上进行打印
28                // execute program
29                env.execute("Flink Streaming Java API Skeleton");
30
31        }
32
33import org.apache.flink.streaming.api.functions.AssignerWithPunctuatedWatermarks;
34import org.apache.flink.streaming.api.watermark.Watermark;
35
36public class CustomWatermarkEmitter implements AssignerWithPunctuatedWatermarks<String> {
37
38    private static final long serialVersionUID = 1L;
39
40    public long extractTimestamp(String arg0, long arg1) {
41        if (null != arg0 && arg0.contains(",")) {
42            String parts[] = arg0.split(",");
43            return Long.parseLong(parts[0]);
44        }
45        return 0;
46    }
47
48    public Watermark checkAndGetNextWatermark(String arg0, long arg1) {
49        if (null != arg0 && arg0.contains(",")) {
50            String parts[] = arg0.split(",");
51            return new Watermark(Long.parseLong(parts[0]));
52        }
53        return null;
54    }
55}

4.开启一台配置好zookeeper和kafka的Ubuntu虚拟机，输入以下命令分别开启zookeeper、kafka、topic、producer。(zookeeper和kafka的配置可参考https://www.cnblogs.com/ALittleMoreLove/p/9396745.html)

1bin/zkServer.sh start
2bin/kafka-server-start.sh config/server.properties
3bin/kafka-topics.sh --create --zookeeper 10.192.12.106:2181 --replication-factor 1 --partitions 1 --topic test0
4bin/kafka-console-producer.sh --broker-list 10.192.12.106:9092 --topic test0

5.检测Flink程序是否可以接收到来自Kafka生产者发来的数据，运行Java类KafkaDemo，在开启kafka生产者的终端下随便输入一段话，在IDEA控制台可以收到该信息，如下为kafka生产者终端和控制台。

OK，成功的接收到了来自Kafka生产者的消息^.^。