Storm流之GlobalGrouping全局分组

一、概述

       全局分组方式将所有的 tuples 路由到唯一一个 task 上。Storm选取最小的 task ID 来选取接收数据的 task。注意，当使用全局分组时，设置 bolt 的 task 并发度是没有意义的，因为所有 tuple 都转发到同一个 task 上了。使用全局分组的时候需要注意，因为所有的tuple都转发到一个JVM实例上，可能会引起Storm集群中某个JVM或者服务器出现性能瓶颈或崩溃。

二、代码

1. Spout

package com.test.csdn.globalgrouping;

import org.apache.storm.spout.SpoutOutputCollector;
import org.apache.storm.task.TopologyContext;
import org.apache.storm.topology.OutputFieldsDeclarer;
import org.apache.storm.topology.base.BaseRichSpout;
import org.apache.storm.tuple.Fields;
import org.apache.storm.tuple.Values;

import java.util.Map;
import java.util.Random;

/**
 *
 * @author Simon
 * @date 2018/8/1
 */
public class GlobalGroupingSpout extends BaseRichSpout {

    private SpoutOutputCollector collector;

    private String[] str =  {"xiaomi","huawei","apple","oppo","vivo","lenovo","LG",
            "samsung","htc","honor","nokia","smartisan","Sony","BlackBerry","sharp"};


    @Override
    public void open(Map map, TopologyContext topologyContext, SpoutOutputCollector spoutOutputCollector) {
        this.collector=spoutOutputCollector;
    }

    @Override
    public void nextTuple() {
        try {
            Thread.sleep(1000);
            int i = new Random().nextInt(10);
            String string = str[i];
            collector.emit(new Values(string));
        } catch (InterruptedException e) {
            e.printStackTrace();
        }

    }

    @Override
    public void declareOutputFields(OutputFieldsDeclarer outputFieldsDeclarer) {
        outputFieldsDeclarer.declare(new Fields("string"));
    }
}

    2.Bolt

package com.test.csdn.globalgrouping;

import org.apache.storm.topology.BasicOutputCollector;
import org.apache.storm.topology.OutputFieldsDeclarer;
import org.apache.storm.topology.base.BaseBasicBolt;

import org.apache.storm.tuple.Tuple;


/**
 * Created by Simon on 2018/8/1.
 */
public class GlobalGroupingBolt extends BaseBasicBolt {

    @Override
    public void execute(Tuple tuple, BasicOutputCollector basicOutputCollector) {
        System.out.println(Thread.currentThread().getName()+"___"+tuple.getValue(0));
    }

    @Override
    public void declareOutputFields(OutputFieldsDeclarer outputFieldsDeclarer) {

    }
}

       3.Topo

package com.test.csdn.globalgrouping;

import com.test.csdn.fieldsgrouping.FiledGroupingBolt;
import com.test.csdn.fieldsgrouping.FiledGroupingSpout;
import org.apache.storm.Config;
import org.apache.storm.LocalCluster;
import org.apache.storm.topology.TopologyBuilder;
import org.apache.storm.tuple.Fields;
import org.apache.storm.utils.Utils;

/**
 * Created by Simon on 2018/8/1.
 */
public class GlobalGroupingTopo {
    public static void main(String[] args) {

        TopologyBuilder builder = new TopologyBuilder();

        builder.setSpout("spout", new GlobalGroupingSpout()).setNumTasks(3);
        builder.setBolt("bolt", new GlobalGroupingBolt()).setNumTasks(5).globalGrouping("spout");

        Config conf = new Config();
        conf.setDebug(false);

        LocalCluster cluster = new LocalCluster();
        cluster.submitTopology("toplogy", conf, builder.createTopology());
        Utils.sleep(Long.MAX_VALUE);
        cluster.shutdown();

    }
}

 三、运行输出

        

           运行在同一个taskId里面。