HBase 之TableOutputFormat

    MapReducer的输出导入到HBase有多种方式可以实现， TableOutputFormat就是其中一种.

1. hbase建表

hbase(main):132:0* create 't1','f1'
0 row(s) in 1.4890 seconds

hbase(main):133:0> scan 't1'
ROW                              COLUMN+CELL
0 row(s) in 1.2330 seconds

2.写MR作业
HBaseMapper.java

public class HBaseMapper extends MapReduceBase implements Mapper<LongWritable, Text, LongWritable, Text> {
    @Override
    public void map(LongWritable key, Text values,
            OutputCollector<LongWritable, Text> output, Reporter reporter)
            throws IOException {
        output.collect(key, values);
    }
}

HBaseReducer.java

public class HBaseReducer extends MapReduceBase implements Reducer<LongWritable, Text, ImmutableBytesWritable, Put> {
    @Override
    public void reduce(LongWritable key, Iterator<Text> values,
            OutputCollector<ImmutableBytesWritable, Put> output, Reporter reporter)
            throws IOException {
        String value="";
        ImmutableBytesWritable immutableBytesWritable = new ImmutableBytesWritable();
        Text text = new Text();
        while(values.hasNext())
        {
            value = values.next().toString();
            if(value != null && !"".equals(value))
            {
                Put put = createPut(value.toString());
                if(put!=null)
                    output.collect(immutableBytesWritable, put);
            }
        }
    }

// str格式为row:family:qualifier:value 简单模拟下而已
    private Put createPut(String str)
    {
        String[] strstrs = str.split(":");
        if(strs.length<4)
            return null;
        String row=strs[0];
        String family=strs[1];
        String qualifier=strs[2];
        String value=strs[3];
        Put put = new Put(Bytes.toBytes(row));
        put.add(Bytes.toBytes(family), Bytes.toBytes(qualifier), 1L,Bytes.toBytes(value));
        return put;
    }
}

HbaseDriver.java

public class HbaseDriver {
    public static void main(String[] args) {
        JobConf conf = new JobConf(com.test.HbaseDriver.class);
        conf.setMapperClass(com.test.HBaseMapper.class);
        conf.setReducerClass(com.test.HBaseReducer.class);

        conf.setMapOutputKeyClass(LongWritable.class);
        conf.setMapOutputValueClass(Text.class);

        conf.setOutputKeyClass(ImmutableBytesWritable.class);
        conf.setOutputValueClass(Put.class);

        conf.setOutputFormat(TableOutputFormat.class);

        FileInputFormat.setInputPaths(conf, "/home/yinjie/input");
        FileOutputFormat.setOutputPath(conf, new Path("/home/yinjie/output"));

        conf.set(TableOutputFormat.OUTPUT_TABLE, "t1");
        conf.set("hbase.zookeeper.quorum", "localhost");
        conf.set("hbase.zookeeper.property.clientPort", "2181");
        try {
            JobClient.runJob(conf);
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}

/home/yinjie/input目录下有一个hbasedata.txt文件,内容为

[root@localhost input]# cat hbasedata.txt
r1:f1:c1:value1
r2:f1:c2:value2
r3:f1:c3:value3

在eclipse下使用MR插件，运行作业:
作业成功后再次查询hbase表，验证数据是否已进去:

hbase(main):135:0> scan 't1'
ROW                              COLUMN+CELL
 r1                              column=f1:c1, timestamp=1, value=value1
 r2                              column=f1:c2, timestamp=1, value=value2
 r3                              column=f1:c3, timestamp=1, value=value3
3 row(s) in 0.0580 seconds

数据已进插入^_^, TableOutputFormat效率并不好，大数据量装载到hbase的话最好生成HFile后再倒入到hbase, HFile是hbase内部存储表示形式， 所以装载数度很快.