Oracle删除重复记录

经过测试，最高效的一种：(1.5万条数据，大约0.125s)
delete from tableName le　where le.rowid not in
( select MAX(a.rowid) from tableName a GROUP BY a.字段1, a.字段2,... )


上次后台导员工数据，耗费了大量的时间，甚至页面报错，分析原因，在于删除重复记录
的SQL语句太耗时间，经小李的指点，将原有sql进行了优化，优化结果：

---------------------------------------------------------------------------------------

可以将：

delete from netb_employ le　where le.rowid != 　 (select b.dataid
from (select a.num,
a.name,
MAX(a.ROWID) as dataid
from netb_employ a
GROUP BY a.num,
a.name) b　
where le.num = b.num
and le.name = b.name) ;

耗时：大约25秒

---------------------------------------------------------------------------------------

优化为：


delete from netb_employ le　where le.rowid not in 　
( select MAX(a.ROWID) from netb_employ a GROUP BY a.num, a.name ) ;

耗时：不到半秒
----------------------------------------------------------------------------------------