1) EXPDP\IMPDP
采用逻辑导出导入很简单，首先在源库建立分区表，然后将数据导出，然后导入到新建的分区表即可

2) Insert into
利用原表重建分区表,再insert into 分区表 select *　from 原表

3) Exchange Partition(交换分区)

4) DBMS_REDEFINITION(在线重定义)

简单介绍1、2两种的实验过程
普通表数据如下
conn hr/hr
SQL> select * from CUSTOMER99;
CUST_NAME CUST_ID
---------------------------------------- ----------
1 1
11 11
21 21
31 31
41 41

--Insert into的案例
SQL>CREATE TABLE customer199
( cust_name varchar2(20)
, cust_id NUMBER
)
PARTITION BY RANGE (cust_id)
( PARTITION q1t VALUES LESS THAN (10)
, PARTITION q2t VALUES LESS THAN (20)
, PARTITION q3t VALUES LESS THAN (30)
, PARTITION q4t VALUES LESS THAN (40)
, partition q5t VALUES LESS THAN(maxvalue)
);

SQL>insert into customer199 select * from CUSTOMER99;

SQL> select * from customer199 partition(q3t);
CUST_NAME CUST_ID
---------------------------------------- ----------
21 21

--EXPDP\IMPDP的案例
expdp hr/hr directory=DATA_PUMP_DIR dumpfile=CUSTOMER99.dmp tables=CUSTOMER99
CREATE TABLE customer_expdp
( cust_name varchar2(20)
, cust_id NUMBER
)
PARTITION BY RANGE (cust_id)
( PARTITION q11 VALUES LESS THAN (10)
, PARTITION q21 VALUES LESS THAN (20)
, PARTITION q31 VALUES LESS THAN (30)
, PARTITION q41 VALUES LESS THAN (40)
, partition q51 VALUES LESS THAN(maxvalue)
);
impdp hr/hr directory=DATA_PUMP_DIR dumpfile=CUSTOMER99.dmp remap_table=CUSTOMER99:customer_expdp table_exists_action=append
SQL> select * from customer_expdp partition(q31);
CUST_NAME CUST_ID
---------------------------------------- ----------
21 21

ITPub博客

普通表转换成分区表的四种方法

最新文章