change Reduce_Number Of Order_By

[郭帅]

I have some problems when I insert data with order by .
The data is non-uniform ， I changed  the configuration below 
     1. set hive.optimize.sampling.orderby=true;
     2. set hive.optimize.sampling.orderby.number=100000;
     3.set mapred.reduce.tasks=100;
but when the sampling had done , there were only 2 reducer for the job . 
Could I  make it 100   and How.


--
 have a nice day～guoshuai