【HBase基础教程】6、HBase之读取MapReduce数据写入HBase

本blog将介绍利用MapReduce操作HBase，借助最熟悉的单词计数案例WordCount，将WordCount的统计结果存储到HBase，而不是HDFS。

开发环境

硬件环境：Centos 6.5 服务器4台（一台为Master节点，三台为Slave节点）
软件环境：Java 1.7.0_45、Eclipse Juno Service Release 2、hadoop-1.2.1、hbase-0.94.20。

1、 输入与输出

1）输入文件

file0.txt（WordCountHbaseWriter\input\file0.txt）    
Hello World Bye World   
file1.txt（WordCountHbaseWriter\input\file1.txt）   
Hello Hadoop Goodbye Hadoop   

2）输出HBase数据库

以下为输出数据库wordcount的数据库结构，以及预期的输出结果，如下图所示：

2、 Mapper函数实现

WordCountHbaseMapper程序和WordCount的Map程序一样，Map输入为每一行数据，例如”Hello World Bye World”，通过StringTokenizer类按空格分割成一个个单词，
通过context.write(word, one);输出为一系列< key,value>键值对：<”Hello”,1><”World”,1><”Bye”,1><”World”,1>。
详细源码请参考：WordCountHbaseWriter\src\com\zonesion\hbase\WordCountHbaseWriter.java

public static class WordCountHbaseMapper extends
        Mapper<Object, Text, Text, IntWritable> {
    private final static IntWritable one = new IntWritable(1);
    private Text word = new Text();

    public void map(Object key, Text value, Context context)
            throws IOException, InterruptedException {
        StringTokenizer itr = new StringTokenizer(value.toString());
        while (itr.hasMoreTokens()) {
            word.set(itr.nextToken());
            context.write(word, one);// 输出<key,value>为<word,one>
        }
    }
}

3、 Reducer函数实现

WordCountHbaseReducer继承的是TableReducer类，在Hadoop中TableReducer继承Reducer类，它的原型为TableReducer< KeyIn,Values,KeyOut>，前两个参数必须对应Map过程的输出类型key/value类型，第三个参数为ImmutableBytesWritable，即为不可变类型。reduce(Text key, Iterable< IntWritable> values,Context context)具体处理过程分析如下表所示。

详细源码请参考：WordCountHbaseWriter\src\com\zonesion\hbase\WordCountHbaseWriter.java

public static class WordCountHbaseReducer extends
            TableReducer<Text, IntWritable, ImmutableBytesWritable> {

        public void reduce(Text key, Iterable<IntWritable> values,
                Context context) throws IOException, InterruptedException {
            int sum = 0;
            for (IntWritable val : values) {// 遍历求和
                sum += val.get();
            }
            Put put = new Put(key.getBytes());//put实例化，每一个词存一行
            //列族为content,列修饰符为count，列值为数目
            put.add(Bytes.toBytes("content"), Bytes.toBytes("count"), Bytes.toBytes(String.valueOf(sum)));
            context.write(new ImmutableBytesWritable(key.getBytes()), put);// 输出求和后的<key,value>
        }
    }

4、 驱动函数实现

与WordCount的驱动类不同，在Job配置的时候没有配置job.setReduceClass()，而是用以下方法执行Reduce类：

TableMapReduceUtil.initTableReducerJob(tablename, WordCountHbaseReducer.class, job);

该方法指明了在执行job的reduce过程时，执行WordCountHbaseReducer，并将reduce的结果写入到表明为tablename的表中。特别注意：此处的TableMapReduceUtil是hadoop.hbase.mapreduce包中的，而不是hadoop.hbase.mapred包中的，否则会报错。
详细源码请参考：WordCountHbaseWriter\src\com\zonesion\hbase\WordCountHbaseWriter.java

public static void main(String[] args) throws Exception {
    String tablename = "wordcount";
    Configuration conf = HBaseConfiguration.create();
    conf.set("hbase.zookeeper.quorum", "Master");
    HBaseAdmin admin = new HBaseAdmin(conf);
    if(admin.tableExists(tablename)){
        System.out.println("table exists!recreating.......");
        admin.disableTable(tablename);
        admin.deleteTable(tablename);
    }
    HTableDescriptor htd = new HTableDescriptor(tablename);
    HColumnDescriptor tcd = new HColumnDescriptor("content");
    htd.addFamily(tcd);//创建列族
    admin.createTable(htd);//创建表
    String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();
    if (otherArgs.length != 1) {
      System.err.println("Usage: WordCountHbase <in>");
      System.exit(2);
    }
    Job job = new Job(conf, "WordCountHbase");
    job.setJarByClass(WordCountHbase.class);
    //使用WordCountHbaseMapper类完成Map过程；
    job.setMapperClass(WordCountHbaseMapper.class);
    TableMapReduceUtil.initTableReducerJob(tablename, WordCountHbaseReducer.class, job);
    //设置任务数据的输入路径；
    FileInputFormat.addInputPath(job, new Path(otherArgs[0]));
    //设置了Map过程和Reduce过程的输出类型，其中设置key的输出类型为Text；
    job.setOutputKeyClass(Text.class);
    //设置了Map过程和Reduce过程的输出类型，其中设置value的输出类型为IntWritable；
    job.setOutputValueClass(IntWritable.class);
    //调用job.waitForCompletion(true) 执行任务，执行成功后退出；
    System.exit(job.waitForCompletion(true) ? 0 : 1);
}

5、部署运行

1）启动Hadoop集群和Hbase服务

[hadoop@K-Master ~]$ start-dfs.sh     #启动hadoop HDFS文件管理系统
[hadoop@K-Master ~]$ start-mapred.sh      #启动hadoop MapReduce分布式计算服务
[hadoop@K-Master ~]$ start-hbase.sh       #启动Hbase
[hadoop@K-Master ~]$ jps              #查看进程
22003 HMaster
10611 SecondaryNameNode
22226 Jps
21938 HQuorumPeer
10709 JobTracker
22154 HRegionServer
20277 Main
10432 NameNode

特别注意：用户可先通过jps命令查看Hadoop集群和Hbase服务是否启动，如果Hadoop集群和Hbase服务已经启动，则不需要执行此操作。

2）部署源码

#设置工作环境
[hadoop@K-Master ~]$ mkdir -p /usr/hadoop/workspace/Hbase
#部署源码
将WordCountHbaseWriter文件夹拷贝到/usr/hadoop/workspace/Hbase/ 路径下；

… 你可以直接 下载 WordCountHbaseWriter

3）修改配置文件

a) 查看hbase核心配置文件hbase-site.xml的hbase.zookeeper.quorum属性

参考“【HBase基础教程】5、HBase API访问 3、部署运行 3）修改配置文件”查看hbase核心配置文件hbase-site.xml的hbase.zookeeper.quorum属性；

b) 修改项目WordCountHbaseWriter/src/config.properties属性文件

将项目WordCountHbaseWriter/src/config.properties属性文件的hbase.zookeeper.quorum属性值修改为上一步查询到的属性值，保持config.properties文件的hbase.zookeeper.quorum属性值与hbase-site.xml文件的hbase.zookeeper.quorum属性值一致；

#切换工作目录
[hadoop@K-Master ~]$ cd /usr/hadoop/workspace/Hbase/WordCountHbaseWriter
#修改属性值
[hadoop@K-Master WordCountHbaseWriter]$ vim src/config.properties
hbase.zookeeper.quorum=K-Master
#拷贝src/config.properties文件到bin/文件夹
[hadoop@K-Master WordCountHbaseWriter]$ cp src/config.properties bin/

4）上传输入文件

#创建输入文件夹
[hadoop@K-Master WordCountHbaseWriter]$ hadoop fs -mkdir HbaseWriter/input/
#上传文件到输入文件夹 
[hadoop@K-Master WordCountHbaseWriter]$ hadoop fs -put input/file* HbaseWriter/input/
#查看上传文件是否成功
[hadoop@K-Master WordCountHbaseWriter]$ hadoop fs -ls HbaseWriter/input/
Found 2 items
-rw-r--r--   3 hadoop supergroup 22 2014-12-30 17:39 /user/hadoop/HbaseWriter/input/file0.txt
-rw-r--r--   3 hadoop supergroup 28 2014-12-30 17:39 /user/hadoop/HbaseWriter/input/file1.txt

5）编译文件

#执行编译
[hadoop@K-Master WordCountHbaseWriter]$ javac -classpath /usr/hadoop/hadoop-core-1.2.1.jar:/usr/hadoop/lib/commons-cli-1.2.jar:lib/zookeeper-3.4.5.jar:lib/hbase-0.94.20.jar -d bin/ src/com/zonesion/hbase/*.java
#查看编译是否成功
[hadoop@K-Master WordCountHbaseWriter]$ ls bin/com/zonesion/hbase/ -la
total 24
drwxrwxr-x 2 hadoop hadoop 4096 Dec 30 17:20 .
drwxrwxr-x 3 hadoop hadoop 4096 Dec 30 17:20 ..
-rw-rw-r-- 1 hadoop hadoop 3446 Dec 30 17:29 PropertiesHelper.class
-rw-rw-r-- 1 hadoop hadoop 3346 Dec 30 17:29 WordCountHbaseWriter.class
-rw-rw-r-- 1 hadoop hadoop 1817 Dec 30 17:29 WordCountHbaseWriter$WordCountHbaseMapper.class
-rw-rw-r-- 1 hadoop hadoop 2217 Dec 30 17:29 WordCountHbaseWriter$WordCountHbaseReducer.class

6） 打包Jar文件

#拷贝lib文件夹到bin文件夹
[hadoop@K-Master WordCountHbaseWriter]$ cp –r lib/ bin/
#打包Jar文件
[hadoop@K-Master WordCountHbaseWriter]$ jar -cvf WordCountHbaseWriter.jar -C bin/ . 
added manifest
adding: lib/(in = 0) (out= 0)(stored 0%)
adding: lib/zookeeper-3.4.5.jar(in = 779974) (out= 721150)(deflated 7%)
adding: lib/guava-11.0.2.jar(in = 1648200) (out= 1465342)(deflated 11%)
adding: lib/protobuf-java-2.4.0a.jar(in = 449818) (out= 420864)(deflated 6%)
adding: lib/hbase-0.94.20.jar(in = 5475284) (out= 5038635)(deflated 7%)
adding: com/(in = 0) (out= 0)(stored 0%)
adding: com/zonesion/(in = 0) (out= 0)(stored 0%)
adding: com/zonesion/hbase/(in = 0) (out= 0)(stored 0%)
adding: com/zonesion/hbase/WordCountHbaseWriter.class(in = 3136) (out= 1583)(deflated 49%)
adding: com/zonesion/hbase/WordCountHbaseWriter$WordCountHbaseMapper.class(in = 1817) (out= 772)(deflated 57%)
adding: com/zonesion/hbase/WordCountHbaseWriter$WordCountHbaseReducer.class(in = 2217) (out= 929)(deflated 58%)

7）运行实例

[hadoop@K-Master WordCountHbaseWriter]$ hadoop jar WordCountHbaseWriter.jar com.zonesion.hbase.WordCountHbaseWriter /user/hadoop/HbaseWriter/input/
...................省略.............
14/12/30 11:23:59 INFO input.FileInputFormat: Total input paths to process : 2
14/12/30 11:23:59 INFO util.NativeCodeLoader: Loaded the native-hadoop library
14/12/30 11:23:59 WARN snappy.LoadSnappy: Snappy native library not loaded
14/12/30 11:24:05 INFO mapred.JobClient: Running job: job_201412161748_0020
14/12/30 11:24:06 INFO mapred.JobClient:  map 0% reduce 0%
14/12/30 11:24:27 INFO mapred.JobClient:  map 50% reduce 0%
14/12/30 11:24:30 INFO mapred.JobClient:  map 100% reduce 0%
14/12/30 11:24:39 INFO mapred.JobClient:  map 100% reduce 100%
14/12/30 11:24:41 INFO mapred.JobClient: Job complete: job_201412161748_0020
14/12/30 11:24:41 INFO mapred.JobClient: Counters: 28
14/12/30 11:24:41 INFO mapred.JobClient:   Job Counters
14/12/30 11:24:41 INFO mapred.JobClient: Launched reduce tasks=1
14/12/30 11:24:41 INFO mapred.JobClient: SLOTS_MILLIS_MAPS=20955
14/12/30 11:24:41 INFO mapred.JobClient: Total time spent by all reduces waiting after reserving slots (ms)=0
14/12/30 11:24:41 INFO mapred.JobClient: Total time spent by all maps waiting after reserving slots (ms)=0
14/12/30 11:24:41 INFO mapred.JobClient: Launched map tasks=2
14/12/30 11:24:41 INFO mapred.JobClient: Data-local map tasks=2
14/12/30 11:24:41 INFO mapred.JobClient: SLOTS_MILLIS_REDUCES=11527
14/12/30 11:24:41 INFO mapred.JobClient:   File Output Format Counters
14/12/30 11:24:41 INFO mapred.JobClient: Bytes Written=0
14/12/30 11:24:41 INFO mapred.JobClient:   FileSystemCounters
14/12/30 11:24:41 INFO mapred.JobClient: FILE_BYTES_READ=104
14/12/30 11:24:41 INFO mapred.JobClient: HDFS_BYTES_READ=296
14/12/30 11:24:41 INFO mapred.JobClient: FILE_BYTES_WRITTEN=239816
14/12/30 11:24:41 INFO mapred.JobClient:   File Input Format Counters
14/12/30 11:24:41 INFO mapred.JobClient: Bytes Read=50
14/12/30 11:24:41 INFO mapred.JobClient:   Map-Reduce Framework
14/12/30 11:24:41 INFO mapred.JobClient: Map output materialized bytes=110
14/12/30 11:24:41 INFO mapred.JobClient: Map input records=2
14/12/30 11:24:41 INFO mapred.JobClient: Reduce shuffle bytes=110
14/12/30 11:24:41 INFO mapred.JobClient: Spilled Records=16
14/12/30 11:24:41 INFO mapred.JobClient: Map output bytes=82
14/12/30 11:24:41 INFO mapred.JobClient: Total committed heap usage (bytes)=417546240
14/12/30 11:24:41 INFO mapred.JobClient: CPU time spent (ms)=1110
14/12/30 11:24:41 INFO mapred.JobClient: Combine input records=0
14/12/30 11:24:41 INFO mapred.JobClient: SPLIT_RAW_BYTES=246
14/12/30 11:24:41 INFO mapred.JobClient: Reduce input records=8
14/12/30 11:24:41 INFO mapred.JobClient: Reduce input groups=5
14/12/30 11:24:41 INFO mapred.JobClient: Combine output records=0
14/12/30 11:24:41 INFO mapred.JobClient: Physical memory (bytes) snapshot=434167808
14/12/30 11:24:41 INFO mapred.JobClient: Reduce output records=5
14/12/30 11:24:41 INFO mapred.JobClient: Virtual memory (bytes) snapshot=2192027648
14/12/30 11:24:41 INFO mapred.JobClient: Map output records=8

8）查看输出结果

#另外开启一个终端，输入hbase shell命令进入hbase shell命令行
[hadoop@K-Master ~]$ hbase shell
HBase Shell; enter 'help<RETURN>' for list of supported commands.
Type "exit<RETURN>" to leave the HBase Shell
Version 0.94.20, r09c60d770f2869ca315910ba0f9a5ee9797b1edc, Fri May 23 22:00:41 PDT 2014

hbase(main):002:0> scan 'wordcount'
ROW   COLUMN+CELL
 Bye  column=content:count, timestamp=1419932527321, value=1
 Goodbye  column=content:count, timestamp=1419932527321, value=1
 Hadoope  column=content:count, timestamp=1419932527321, value=2
 Hellope  column=content:count, timestamp=1419932527321, value=2
 Worldpe  column=content:count, timestamp=1419932527321, value=2
5 row(s) in 0.6370 seconds
 

您可能喜欢

【HBase基础教程】1、HBase之单机模式与伪分布式模式安装
【HBase基础教程】2、HBase之完全分布式模式安装
【HBase基础教程】3、HBase Shell DDL操作
【HBase基础教程】4、HBase Shell DML操作
【HBase基础教程】5、HBase API访问
【HBase基础教程】6、HBase之读取MapReduce数据写入HBase
【HBase基础教程】7、HBase之读取HBase数据写入HDFS