1、首先在HIVE里面创建一个名叫DUAL的表。create table dual (dummy string);
2、创建好了dual表之后可以直接退出hive，回到Linux里面。生成一个dual.txt的文件。echo 'X' > dual.txt
3、接着就是把dual.txt放到hadoop的hdfs里面去，跟dual表关联起来。>hive load data local inpath '/home/hadoop/dual.txt' overwrite into table daul;
4、好了，这个时候可以尝试下是否能实现向oracle里面的dual的功能了。hive>select 'hello' from dual;
5、这是得出来的结果（注意加粗字体部分的就可以了。）：hive> select 'hello' from dual;Total MapRedu艘早祓胂ce jobs = 1Launching Job 1 out of 1Number of reduce tasks is set to 0 since there's no reduce operatorStarting Job = job_201507031544_0002, Tracking URL = http://master:50030/jobdetails.jsp?jobid=job_201507031544_0002Kill Command = /home/james/hadoop/libexec/../bin/hadoop job -kill job_201507031544_0002Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 02015-07-30 19:13:52,669 Stage-1 map = 0%, reduce = 0%2015-07-30 19:13:55,784 Stage-1 map = 100%, reduce = 0%, Cumulative CPU 0.96 sec2015-07-30 19:13:56,849 Stage-1 map = 100%, reduce = 100%, Cumulative CPU 0.96 secMapReduce Total cumulative CPU time: 960 msecEnded Job = job_201507031544_0002MapReduce Jobs Launched:Job 0: Map: 1 Cumulative CPU: 0.96 sec HDFS Read: 217 HDFS Write: 6 SUCCESSTotal MapReduce CPU Time Spent: 960 msecOKhelloTime taken: 27.577 seconds, Fetched: 1 row(s)
6、好了，到此Oracle中dual的功能可以在HIVE里面实现了。非常的简单！
