使用hadoop mapreduce分析mongodb数据：（1）

最近考虑使用hadoop mapreduce来分析mongodb上的数据，从网上找了一些demo，东拼西凑，终于运行了一个demo，下面把过程展示给大家

环境

• ubuntu 14.04 64bit
• hadoop 2.6.4
• mongodb 2.4.9
• Java 1.8
• mongo-hadoop-core-1.5.2.jar
• mongo-java-driver-3.0.4.jar

mongo-hadoop-core-1.5.2.jar以及mongo-java-driver-3.0.4.jar的下载和配置

• 编译mongo-hadoop-core-1.5.2.jar

• $ git clone https://github.com/mongodb/mongo-hadoop
  $ cd mongo-hadoop
  $ ./gradlew jar

  • 编译时间比较长，成功编译之后mongo-hadoop-core-1.5.2.jar存在的路径是core/build/libs
• 下载mongo-java-driver-3.0.4.jar

• http://central.maven.org/maven2/org/mongodb/mongo-java-driver/3.0.4/
  选择 mongo-java-driver-3.0.4.jar

数据

• 数据样例

• > db.in.find({})
  { "_id" : ObjectId("5758db95ab12e17a067fbb6f"), "x" : "hello world" }
  { "_id" : ObjectId("5758db95ab12e17a067fbb70"), "x" : "nice to meet you" }
  { "_id" : ObjectId("5758db95ab12e17a067fbb71"), "x" : "good to see you" }
  { "_id" : ObjectId("5758db95ab12e17a067fbb72"), "x" : "world war 2" }
  { "_id" : ObjectId("5758db95ab12e17a067fbb73"), "x" : "see you again" }
  { "_id" : ObjectId("5758db95ab12e17a067fbb74"), "x" : "bye bye" }

• 最后的结果

• > db.out.find({})
  { "_id" : "2", "value" : 1 }
  { "_id" : "again", "value" : 1 }
  { "_id" : "bye", "value" : 2 }
  { "_id" : "good", "value" : 1 }
  { "_id" : "hello", "value" : 1 }
  { "_id" : "meet", "value" : 1 }
  { "_id" : "nice", "value" : 1 }
  { "_id" : "see", "value" : 2 }
  { "_id" : "to", "value" : 2 }
  { "_id" : "war", "value" : 1 }
  { "_id" : "world", "value" : 2 }
  { "_id" : "you", "value" : 3 }

• 目标是统计每个文档中出现的词频，并且把单词作为key，词频作为value存在mongodb中

Hadoop mapreduce代码

• Mapreduce 代码

  import java.util.*; 
  import java.io.*;
  
  import org.bson.*;
  
  import com.mongodb.hadoop.MongoInputFormat;
  import com.mongodb.hadoop.MongoOutputFormat;
  
  import org.apache.hadoop.conf.Configuration;
  import org.apache.hadoop.io.*;
  import org.apache.hadoop.mapreduce.*;
  
  
  public class WordCount {
      public static class TokenizerMapper extends Mapper<Object, BSONObject, Text, IntWritable> {
          private final static IntWritable one = new IntWritable(1);
          private Text word = new Text();
          public void map(Object key, BSONObject value, Context context ) 
                  throws IOException, InterruptedException {
              System.out.println( "key: " + key );
              System.out.println( "value: " + value );
              StringTokenizer itr = new StringTokenizer(value.get( "x" ).toString());
              while (itr.hasMoreTokens()) {
                  word.set(itr.nextToken());
                  context.write(word, one);
              }
          }
      }
      public static class IntSumReducer extends Reducer<Text,IntWritable,Text,IntWritable> {
          private IntWritable result = new IntWritable();
          public void reduce(Text key, Iterable<IntWritable> values, Context context )
              throws IOException, InterruptedException {
              int sum = 0;
              for (IntWritable val : values) {
                  sum += val.get();
              }
              result.set(sum);
              context.write(key, result);
          }
      }
      public static void main(String[] args) throws Exception {
          Configuration conf = new Configuration();
          conf.set( "mongo.input.uri" , "mongodb://localhost/testmr.in" );
          conf.set( "mongo.output.uri" , "mongodb://localhost/testmr.out" );
          @SuppressWarnings("deprecation")
          Job job = new Job(conf, "word count");
          job.setJarByClass(WordCount.class);
          job.setMapperClass(TokenizerMapper.class);
          job.setCombinerClass(IntSumReducer.class);
          job.setReducerClass(IntSumReducer.class);
          job.setOutputKeyClass(Text.class);
          job.setOutputValueClass(IntWritable.class);
          job.setInputFormatClass( MongoInputFormat.class );
          job.setOutputFormatClass( MongoOutputFormat.class );
          System.exit(job.waitForCompletion(true) ? 0 : 1);
      }
  }

• 注意：设置mongo.input.uri和mongo.output.uri

  conf.set( "mongo.input.uri" , "mongodb://localhost/testmr.in" );
  conf.set( "mongo.output.uri" , "mongodb://localhost/testmr.out" );

• 编译
  • 编译

    $ hadoop com.sun.tools.javac.Main WordCount.java -Xlint:deprecation

  • 编译jar包

    $ jar cf wc.jar WordCount*.class

• 运行
  • 启动hadoop，运行mapreduce代码必须启动hadoop

    $ start-all.sh

  • 运行程序

  • $ hadoop jar  wc.jar WordCount

• 查看结果

• $ mongo
  MongoDB shell version: 2.4.9
  connecting to: test
  > use testmr;
  switched to db testmr
  > db.out.find({})
  { "_id" : "2", "value" : 1 }
  { "_id" : "again", "value" : 1 }
  { "_id" : "bye", "value" : 2 }
  { "_id" : "good", "value" : 1 }
  { "_id" : "hello", "value" : 1 }
  { "_id" : "meet", "value" : 1 }
  { "_id" : "nice", "value" : 1 }
  { "_id" : "see", "value" : 2 }
  { "_id" : "to", "value" : 2 }
  { "_id" : "war", "value" : 1 }
  { "_id" : "world", "value" : 2 }
  { "_id" : "you", "value" : 3 }
  > 

以上是一个简单的例子，接下来我要用hadoop mapreduce处理mongodb中的更加复杂的数据。敬请期待，如果有疑问，请在留言区提出 ^_^

参考资料以及文档

1. The elephant in the room mongo db + hadoop
2. http://chenhua-1984.iteye.com/blog/2162576
3. http://api.mongodb.com/java/2.12/com/mongodb/MongoURI.html
4. http://stackoverflow.com/questions/27020075/mongo-hadoop-connector-issue

如果The elephant in the room mongo db + hadoop打不开，请到我的github下载ppt