开发Flink本地WordCount程序（Java）问题总结

在刚接触Flink时，踩了很多坑，接下来就把自己的遇到的问题和大家分享：

首先分享正确的过程：

1、在idea中新建maven项目，在pom.xml文件中导入依赖

        <dependency>
            <groupId>org.apache.flink</groupId>
            <artifactId>flink-streaming-java_2.11</artifactId>
            <version>1.6.4</version>
            <scope>provided</scope>
        </dependency>

2、编写java程序

package TestFlink;

import org.apache.flink.api.common.functions.FlatMapFunction;
import org.apache.flink.api.java.tuple.Tuple;
import org.apache.flink.api.java.tuple.Tuple2;
import org.apache.flink.streaming.api.datastream.DataStream;
import org.apache.flink.streaming.api.datastream.DataStreamSource;
import org.apache.flink.streaming.api.datastream.KeyedStream;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.util.Collector;

public class flinkWordCount {
    public static void main(String[] args) throws Exception{
        //1、初始化flink的环境
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        //2、DataSource
        DataStreamSource<String> source = env.socketTextStream("localhost", 5001);
        //3、DataProcess
        DataStream<Tuple2<String, Integer>> wordones = source.flatMap(new FlatMapFunction<String, Tuple2<String, Integer>>() {
            /**
             * @param line (输入字符串)
             * @param out （输出）
             * @throws Exception
             */
            public void flatMap(String line, Collector<Tuple2<String, Integer>> out) throws Exception {
                String[] words = line.split(" ");
                for (String word : words) {
                    //将分割的字符串以二元数组的形式存储
                    Tuple2<String, Integer> wordone = new Tuple2<String, Integer>(word, 1);
                    //将单词组以二元数组输出
                    out.collect(wordone);
                }

            }
        });
        //3.1按照单词进行分组
        KeyedStream<Tuple2<String, Integer>, Tuple> tuple2TupleKeyedStream = wordones.keyBy(0);
        //3.2计算单词出现的总次数
        DataStream<Tuple2<String, Integer>> sum = tuple2TupleKeyedStream.sum(1);
        //4、DataSink(将结果直接打印 )
        sum.print();
        //5、启动并执行程序
        env.execute("flink WordCount");
    }
}

3、启动NC模拟实时数据（打开CMD ，然后输入nc -l - p 5001）

4、启动Java程序，并在CMD中输入字符串 this is a word

接下来分享遇到的问题：

1、报Flink java.lang.NoClassDefFoundError: org/apache/flink/api/common/functions/FlatMapFunction

解决方法，可参考https://blog.csdn.net/walykyy/article/details/105910155

2、报java.lang.IlleagalStateException: No ExecutorFactory found to execute the application

解决方法：修改pom.xml文件中的flink的引用版本，可直接复制上小节正确流程的pom.xml文件依赖。