1.首先给大数据表创建rownum序列号
--增加序列号字段 alter table TEST add xlh number; --填充序列号 update TEST set xlh = rownum;
2.通过xlh字段进行分数据到不同的表中(以建表的方式进行处理)
create table hik_1001 as select clm1 as hik_clm1,clm2 as hik_clm2,clm3 as hik_clm3 from TEST where xlh>=1 and xlh <=5000000; create table hik_1002 as select clm1 as hik_clm1,clm2 as hik_clm2,clm3 as hik_clm3 from TEST where xlh>=5000001 and xlh <=10000000; create table hik_1003 as select clm1 as hik_clm1,clm2 as hik_clm2,clm3 as hik_clm3 from TEST where xlh>=10000001 and xlh <=15000000; create table hik_1004 as select clm1 as hik_clm1,clm2 as hik_clm2,clm3 as hik_clm3 from TEST where xlh>=15000001;
3.给创建的表分别建立主键或者索引
create index hik_1001_index on HIK_1001 (HIK_CLM2) alter table HIK_1001 add constraint HIK_1001_PKEY primary key (HIK_CLM1)
如果数据有重复的话,只需要建立索引就可以.
4.然后对分表后的数据进行处理
原文：http://www.cnblogs.com/fxust/p/7221018.html