strom整合kafka关键就是以strom中的spout当作kafka的消费者来接收生产者传入的数据。
画一个简单的图:
好了,接下来我们直接上代码!
1,先写一个main方法,作为消费者来接受生产者数据。
package cn.itcast.storm.topology;
import storm.kafka.BrokerHosts;
import storm.kafka.KafkaSpout;
import storm.kafka.SpoutConfig;
import storm.kafka.ZkHosts;
import backtype.storm.Config;
import backtype.storm.LocalCluster;
import backtype.storm.StormSubmitter;
import backtype.storm.spout.SchemeAsMultiScheme;
import backtype.storm.topology.TopologyBuilder;
import backtype.storm.tuple.Fields;
import cn.itcast.storm.bolt.WordSpliter;
import cn.itcast.storm.bolt.WriterBolt;
import cn.itcast.storm.spout.MessageScheme;
public class KafkaTopo {
public static void main(String[] args) throws Exception {
String topic = "kafkaStrom"; //执行要消费的topic
String zkRoot = "/kafka-storm";
String spoutId = "KafkaSpout";
//这里相当于一个消费者,所以不知道topic所在的broker是那台,我们只需要指定zk即可, 从zk中取的元数据查看 topic所在broker,从而拿到topic中生产者的数据
//实际开发生产者一般用flume采集数据,之后文章介绍 flume整合kafka
BrokerHosts brokerHosts = new ZkHosts("weekend01:2181,weekend02:2181,weekend03:2181");
SpoutConfig spoutConfig = new SpoutConfig(brokerHosts, topic, zkRoot, spoutId);
spoutConfig.forceFromStart = true;//从头开始读
spoutConfig.scheme = new SchemeAsMultiScheme(new MessageScheme());
TopologyBuilder builder = new TopologyBuilder();
//设置一个spout用来从kaflka消息队列中读取数据并发送给下一级的bolt组件,此处用的spout组件并非自定义的,而是storm中已经开发好的KafkaSpout
//我们只需要传入strom需要的东西即可,kafka为我们继承了类 kafakaSpout()
builder.setSpout("KafkaSpout", new KafkaSpout(spoutConfig));
builder.setBolt("word-spilter", new WordSpliter()).shuffleGrouping(spoutId);
//产生4个文件,以uuid命名,指定了4个bolt同时处理数据,但是每一行数据只会让同一个bolt处理。
builder.setBolt("writer", new WriterBolt(), 4).fieldsGrouping("word-spilter", new Fields("word"));
Config conf = new Config();
conf.setNumWorkers(4);
conf.setNumAckers(0);
conf.setDebug(false);
LocalCluster cluster = new LocalCluster();
//LocalCluster用来将topology提交到本地模拟器运行,方便开发调试
cluster.submitTopology("WordCount", conf, builder.createTopology());
//提交topology到storm集群中运行
//StormSubmitter.submitTopology("kafkaStrom-topo", conf, builder.createTopology());
}
}
2,实现scheme接口并重写里面方法,主要作用是将生产者送来的数据反序列化,然后指定输出每行信息的字段名称,
然后把这个类放到new KafkaSpout(spoutConfig)方法里面,这是kafka整合好的方法,他相当与storm中的spout组件。
package cn.itcast.storm.spout;
import java.io.UnsupportedEncodingException;
import java.util.List;
import backtype.storm.spout.Scheme;
import backtype.storm.tuple.Fields;
import backtype.storm.tuple.Values;
public class MessageScheme implements Scheme {
private static final long serialVersionUID = 8423372426211017613L;
@Override
public List
3,写一个bolt组件处理拆分spout中发来的数据
package cn.itcast.storm.bolt;
import org.apache.commons.lang.StringUtils;
import backtype.storm.topology.BasicOutputCollector;
import backtype.storm.topology.OutputFieldsDeclarer;
import backtype.storm.topology.base.BaseBasicBolt;
import backtype.storm.tuple.Fields;
import backtype.storm.tuple.Tuple;
import backtype.storm.tuple.Values;
public class WordSpliter extends BaseBasicBolt {
private static final long serialVersionUID = -5653803832498574866L;
@Override
public void execute(Tuple input, BasicOutputCollector collector) {
String line = input.getString(0);
String[] words = line.split(" ");
for (String word : words) {
word = word.trim();
if (StringUtils.isNotBlank(word)) {
word = word.toLowerCase();
collector.emit(new Values(word));
}
}
}
@Override
public void declareOutputFields(OutputFieldsDeclarer declarer) {
declarer.declare(new Fields("word"));
}
}
4,写一个bolt组件将数据放到本地或者hdfs集群中去(这里我们写到了本地,存到hdfs集群的话还得启动7台虚拟机,偷个懒。。)
package cn.itcast.storm.bolt;
import java.io.FileWriter;
import java.io.IOException;
import java.util.Map;
import java.util.UUID;
import backtype.storm.task.TopologyContext;
import backtype.storm.topology.BasicOutputCollector;
import backtype.storm.topology.OutputFieldsDeclarer;
import backtype.storm.topology.base.BaseBasicBolt;
import backtype.storm.tuple.Tuple;
/**
* 将数据写入文件
* @author [email protected]
*
*/
public class WriterBolt extends BaseBasicBolt {
private static final long serialVersionUID = -6586283337287975719L;
private FileWriter writer = null;
@Override
public void prepare(Map stormConf, TopologyContext context) {
try {
writer = new FileWriter("c:\\storm-kafka\\" + "wordcount"+UUID.randomUUID().toString());
} catch (IOException e) {
throw new RuntimeException(e);
}
}
@Override
public void declareOutputFields(OutputFieldsDeclarer declarer) {
}
@Override
public void execute(Tuple input, BasicOutputCollector collector) {
String s = input.getString(0);
try {
writer.write(s);
writer.write("\n");
writer.flush();
} catch (IOException e) {
throw new RuntimeException(e);
}
}
}
到这里代码写完了,接下里我们测试下。
首先保证我们的zk集群处于启动状态,并且上面的kafka也处于启动状态
启动main方法(相当于kafka的消费者)
创建topic
bin/kafka-topics.sh --create --zookeeper weekend01:2181 --replication-factor 3 --partitions 1 --topic mytest
启动生产者代码(之前文章有些)或者在虚拟机上以命令行的模式启动一个生产者也行
命令行模式启动
bin/kafka-console-producer.sh --broker-list weekend01:9092 --topic mytest
然后我们随便输入点东西如:
hello kafka strom
111 222 333
aaa bbb ccc
my name is xxx
最好你去本地目录下查看是否产生消费者处理后的目录(目录是上面bolt组件中定义的)
可以看到该目录下产生了4个文件,因为我们在消费者mian方法中指定了bolt执行的线程数为4个,并以uuid命名为。
谢谢观看,