Flink DataStream创建执行环境的正确方式与细节问题

package com.flink.DataStream.env;

import org.apache.flink.api.common.RuntimeExecutionMode;
import org.apache.flink.api.common.typeinfo.Types;
import org.apache.flink.api.java.tuple.Tuple2;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.configuration.RestOptions;
import org.apache.flink.streaming.api.datastream.DataStreamSource;
import org.apache.flink.streaming.api.datastream.SingleOutputStreamOperator;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.util.Collector;

public class flinkEnvDemo {
    public static void main(String[] args) throws Exception {
        //TODO 创建一个Flink的配置对象
        Configuration configuration = new Configuration();
        //默认是8081,我们改为8082
        configuration.set(RestOptions.BIND_PORT, "8082");
        //TODO 创建Flink的执行环境
        StreamExecutionEnvironment streamExecutionEnvironment = StreamExecutionEnvironment
                //.createLocalEnvironment()   //创建本地环境
                //.createRemoteEnvironment()  //远程环境
                //开发过程中直接使用,他会自动判断是本地集群还是远程环境
                //.getExecutionEnvironment();
                //.getExecutionEnvironment(configuration);
                .createLocalEnvironmentWithWebUI(configuration);  //不启动Flink集群也可以有Web UI
        //TODO 流批一体:代码api是同一套 可以指定为流(默认),也可以指定为批
        //TODO 一般不在代码中写死,提交时,指定参数 —Dexeution.runtime-mode=STREAMING/BATCH
        streamExecutionEnvironment.setRuntimeMode(RuntimeExecutionMode.STREAMING);
        //TODO 创建FLink的source为socket数据源
        DataStreamSource<String> dataStreamSource = streamExecutionEnvironment.socketTextStream("localhost", 8888);
        //TODO 扁平化+转换+分组+聚合
        SingleOutputStreamOperator<Tuple2<String, Integer>> singleOutputStreamOperator = dataStreamSource.flatMap(
                        //使用Lamada表达式实现flatMap接口,当然也可以直接new一个匿名类实现,或者在外部单独定义一个接口实现
                        //泛型第一个是输入类型,第二个是输出类型
                        (String s, Collector<Tuple2<String, Integer>> collector) -> {
                            String[] splitResult = s.split(" ");
                            //循环遍历,将数据转为Tuple类型.spark的rdd算子map: _.map((_,1))
                            for (String word : splitResult) {
                                Tuple2<String, Integer> wordsAndOne = Tuple2.of(word, 1);
                                //使用采集器向下游发送数据
                                collector.collect(wordsAndOne);
                            }
                        })
                .returns(Types.TUPLE(Types.STRING, Types.INT))
                .keyBy(
                        (Tuple2<String, Integer> value) -> {
                            return value.f0;
                        }
                ).sum(1);
        //TODO Sink输出
        singleOutputStreamOperator.print();
        //TODO 执行Flink程序,需要抛异常
        streamExecutionEnvironment.execute("Flink Environment Demo");

        //TODO ......
        /**
         * 默认env.execute() 触发一个Flink Job
         * 一个main方法理论上可以指定多个execute,但是没有什么意义,因为指定到第一个就会阻塞掉
         * 但是Flink 提供了异步执行的方式,一个main方法里面executeAsync()的个数 = 生成的Flink Job数
         * */
        //streamExecutionEnvironment.executeAsync();//异步执行
    }
}

你可能感兴趣的:(#,Flink,flink,大数据)