Storm 1.0.2 - 单词计数案例学习

转自： http://www.cnblogs.com/jonyo/p/5861171.html

单词计数拓扑WordCountTopology实现的基本功能就是不停地读入一个个句子，最后输出每个单词和数目并在终端不断的更新结果，拓扑的数据流如下：

　　 id="iframe_0.24710371619761462" src="data:text/html;charset=utf8,%3Cimg%20id=%22img%22%20src=%22http://wiki.sankuai.com/download/attachments/597748427/WordCountTopology.png?version=1&modificationDate=1473561297000&api=v2&_=5861171%22%20style=%22border:none;max-width:1028px%22%3E%3Cscript%3Ewindow.onload%20=%20function%20()%20%7Bvar%20img%20=%20document.getElementById('img');%20window.parent.postMessage(%7BiframeId:'iframe_0.24710371619761462',width:img.width,height:img.height%7D,%20'http://www.cnblogs.com');%7D%3C/script%3E" frameborder="0" scrolling="no" style="margin: 0px; padding: 0px; border-width: initial; border-style: none; width: 20px; height: 20px;">

• 语句输入Spout:  从数据源不停地读入数据，并生成一个个句子，输出的tuple格式：{"sentence":"hello world"}
• 语句分割Bolt: 将一个句子分割成一个个单词，输出的tuple格式：{"word":"hello"}  {"word":"world"}
• 单词计数Bolt: 保存每个单词出现的次数，每接到上游一个tuple后，将对应的单词加1，并将该单词和次数发送到下游去，输出的tuple格式：{"hello":"1"}  {"world":"3"}
• 结果上报Bolt: 维护一份所有单词计数表，每接到上游一个tuple后，更新表中的计数数据，并在终端将结果打印出来。

　 开发步骤：

    1.环境

• 操作系统：mac os 10.10.3
• JDK: jdk1.8.0_40
• IDE: intellij idea 15.0.3
• Maven: apache-maven-3.0.3

　 2.项目搭建

• 在idea新建一个maven项目工程：storm-learning
• 修改pom.xml文件,加入strom核心的依赖，配置slf4j依赖，方便Log输出

<dependencies>
        <dependency>
            <groupId>org.slf4j</groupId>
            <artifactId>slf4j-api</artifactId>
            <version>1.6.1</version>
        </dependency>

        <dependency>
            <groupId>org.apache.storm</groupId>
            <artifactId>storm-core</artifactId>
            <version>1.0.2</version>
        </dependency>
</dependencies>

　3. Spout和Bolt组件的开发

• SentenceSpout
• SplitSentenceBolt
• WordCountBolt
• ReportBolt

SentenceSpout.java

 1 public class SentenceSpout extends BaseRichSpout{
 2 
 3     private SpoutOutputCollector spoutOutputCollector;
 4 
 5     //为了简单,定义一个静态数据模拟不断的数据流产生
 6     private static final String[] sentences={
 7             "The logic for a realtime application is packaged into a Storm topology",
 8             "A Storm topology is analogous to a MapReduce job",
 9             "One key difference is that a MapReduce job eventually finishes whereas a topology runs forever",
10             " A topology is a graph of spouts and bolts that are connected with stream groupings"
11     };
12 
13     private int index=0;
14 
15     //初始化操作
16     public void open(Map map, TopologyContext topologyContext, SpoutOutputCollector spoutOutputCollector) {
17         this.spoutOutputCollector = spoutOutputCollector;
18     }
19 
20     //核心逻辑
21     public void nextTuple() {
22         spoutOutputCollector.emit(new Values(sentences[index]));
23         ++index;
24         if(index>=sentences.length){
25             index=0;
26         }
27     }
28 
29     //向下游输出
30     public void declareOutputFields(OutputFieldsDeclarer outputFieldsDeclarer) {
31         outputFieldsDeclarer.declare(new Fields("sentences"));
32     }
33 }

SplitSentenceBolt.java

 1 public class SplitSentenceBolt extends BaseRichBolt{
 2 
 3     private OutputCollector outputCollector;
 4 
 5     public void prepare(Map map, TopologyContext topologyContext, OutputCollector outputCollector) {
 6         this.outputCollector = outputCollector;
 7     }
 8 
 9     public void execute(Tuple tuple) {
10         String sentence = tuple.getStringByField("sentences");
11         String[] words = sentence.split(" ");
12         for(String word : words){
13             outputCollector.emit(new Values(word));
14         }
15     }
16 
17     public void declareOutputFields(OutputFieldsDeclarer outputFieldsDeclarer) {
18         outputFieldsDeclarer.declare(new Fields("word"));
19     }
20 }

WordCountBolt.java

 1 public class WordCountBolt extends BaseRichBolt{
 2 
 3     //保存单词计数
 4     private Map<String,Long> wordCount = null;
 5 
 6     private OutputCollector outputCollector;
 7 
 8     public void prepare(Map map, TopologyContext topologyContext, OutputCollector outputCollector) {
 9         this.outputCollector = outputCollector;
10         wordCount = new HashMap<String, Long>();
11     }
12 
13     public void execute(Tuple tuple) {
14         String word = tuple.getStringByField("word");
15         Long count = wordCount.get(word);
16         if(count == null){
17             count = 0L;
18         }
19         ++count;
20         wordCount.put(word,count);
21         outputCollector.emit(new Values(word,count));
22     }
23 
24 
25     public void declareOutputFields(OutputFieldsDeclarer outputFieldsDeclarer) {
26         outputFieldsDeclarer.declare(new Fields("word","count"));
27     }
28 }

ReportBolt.java

 1 public class ReportBolt extends BaseRichBolt {
 2     
 3     private static final Logger log = LoggerFactory.getLogger(ReportBolt.class);
 4 
 5     private Map<String, Long> counts = null;
 6 
 7     public void prepare(Map map, TopologyContext topologyContext, OutputCollector outputCollector) {
 8         counts = new HashMap<String, Long>();
 9     }
10 
11     public void execute(Tuple tuple) {
12         String word = tuple.getStringByField("word");
13         Long count = tuple.getLongByField("count");
14         counts.put(word, count);
15         //打印更新后的结果
16         printReport();
17     }
18 
19     public void declareOutputFields(OutputFieldsDeclarer outputFieldsDeclarer) {
20         //无下游输出,不需要代码
21     }
22 
23     //主要用于将结果打印出来,便于观察
24     private void printReport(){
25         log.info("--------------------------begin-------------------");
26         Set<String> words = counts.keySet();
27         for(String word : words){
28             log.info("@report-bolt@: " + word + " ---> " + counts.get(word));
29         }
30         log.info("--------------------------end---------------------");
31     }
32 }

　4.拓扑配置

• WordCountTopology

 1 public class WordCountTopology {
 2 
 3     private static final Logger log = LoggerFactory.getLogger(WordCountTopology.class);
 4 
 5     //各个组件名字的唯一标识
 6     private final static String SENTENCE_SPOUT_ID = "sentence-spout";
 7     private final static String SPLIT_SENTENCE_BOLT_ID = "split-bolt";
 8     private final static String WORD_COUNT_BOLT_ID = "count-bolt";
 9     private final static String REPORT_BOLT_ID = "report-bolt";
10 
11     //拓扑名称
12     private final static String TOPOLOGY_NAME = "word-count-topology";
13 
14     public static void main(String[] args) {
15 
16         log.info(".........begining.......");
17         //各个组件的实例
18         SentenceSpout sentenceSpout = new SentenceSpout();
19         SplitSentenceBolt splitSentenceBolt = new SplitSentenceBolt();
20         WordCountBolt wordCountBolt = new WordCountBolt();
21         ReportBolt reportBolt = new ReportBolt();
22 
23         //构建一个拓扑Builder
24         TopologyBuilder topologyBuilder = new TopologyBuilder();
25 
26         //配置第一个组件sentenceSpout
27         topologyBuilder.setSpout(SENTENCE_SPOUT_ID, sentenceSpout, 2);
28 
29         //配置第二个组件splitSentenceBolt,上游为sentenceSpout,tuple分组方式为随机分组shuffleGrouping
30         topologyBuilder.setBolt(SPLIT_SENTENCE_BOLT_ID, splitSentenceBolt).shuffleGrouping(SENTENCE_SPOUT_ID);
31 
32         //配置第三个组件wordCountBolt,上游为splitSentenceBolt,tuple分组方式为fieldsGrouping,同一个单词将进入同一个task中(bolt实例)
33         topologyBuilder.setBolt(WORD_COUNT_BOLT_ID, wordCountBolt).fieldsGrouping(SPLIT_SENTENCE_BOLT_ID, new Fields("word"));
34 
35         //配置最后一个组件reportBolt,上游为wordCountBolt,tuple分组方式为globalGrouping,即所有的tuple都进入这一个task中
36         topologyBuilder.setBolt(REPORT_BOLT_ID, reportBolt).globalGrouping(WORD_COUNT_BOLT_ID);
37 
38         Config config = new Config();
39 
40         //建立本地集群,利用LocalCluster,storm在程序启动时会在本地自动建立一个集群,不需要用户自己再搭建,方便本地开发和debug
41         LocalCluster cluster = new LocalCluster();
42 
43         //创建拓扑实例,并提交到本地集群进行运行
44         cluster.submitTopology(TOPOLOGY_NAME, config, topologyBuilder.createTopology());
45     }
46 }

  5.拓扑执行

• 方法一：通过IDEA执行

　　在idea中对代码进行编译compile，然后run；

　　观察控制台输出会发现，storm首先在本地自动建立了运行环境，即启动了zookepeer,接着启动nimbus,supervisor；然后nimbus将提交的topology进行分发到supervisor，supervisor启动woker进程，woker进程里利用Executor来运行topology的组件(spout和bolt);最后在控制台发现不断的输出单词计数的结果。

     zookepeer的连接建立

　  nimbus启动

　  supervisor启动

　  worker启动

     Executor启动执行

     结果输出

• 方法二：通过maven来执行
  • 进入到该项目的主目录下:storm-learning
  • mvn compile 进行代码编译，保证代码编译通过
  • 通过mvn执行程序：

    mvn exec:java -Dexec.mainClass="wordCount.WordCountTopology"

  • 控制台输出的结果跟方法一一致

其他资料：

http://www.cnblogs.com/jonyo/p/5861125.html

http://www.cnblogs.com/jonyo/p/5861835.html