在之前的章节中，讨论过了通过 分区+并行等方式来进行超大的表的切分，通过这种方式能够极大的提高数据的平均分布，但是不是最完美的。

#### $1 dba conn details

function partition_split

sub_partition_name=$4

if [[ $sub_partition_name = 'x' ]]

ksh gen_rowid.sh n1/n1 prdowner subscriber_history x 10

SQL> select count(*)from subscriber_history where rowid between 'AAB4VPAAJAAD7qAAAA' and 'AAB4VPAAJAAD/R/EJA'

COUNT(*)

SQL> select count(*)from subscriber_history where rowid between 'AAB4VPAAJAAD/SAAAA' and 'AAB4VPAAKAABV5/EJA'

COUNT(*)

SQL> select count(*)from subscriber_history where rowid between 'AAB4VPAA4AACQCAAAA' and 'AAB4VPAA5AACHx/EJA';

COUNT(*)