ALTER TABLE…TRUNCATE PARTITION命令用於從指定的分區中刪除資料,只留下完整的分區結構。
文法介紹
文法如下:
ALTER TABLE table_name TRUNCATE PARTITION partition_name
[{DROP|REUSE} STORAGE]
概述
使用ALTER TABLE…TRUNCATE PARTITION命令從指定的分區中刪除資料,只留下完整的分區結構。當您截斷一個分區時,這個分區中的所有子分區也會被截斷。
ALTER TABLE…TRUNCATE PARTITION不會引起可能因表而存在的觸發器ON DELETE起火,但會使觸發器ON TRUNCATE起火。如果為分區定義了觸發器ON TRUNCATE,那麼就會使所有的觸發器BEFORE TRUNCATE在截斷操作前就起火,並使所有的觸發器AFTER TRUNCATE在最後一個截斷操作後也起火。
您必須對一個表有TRUNCATE許可權,才能調用ALTER TABLE…TRUNCATE PARTITION。
參數
參數 | 參數說明 |
table name | 分區表的名稱(可以採用模式限定的方法引用)。 |
partition name | 要刪除的分區名稱。 |
說明 僅在相容情況下包括DROP STORAGE和REUSE STORAGE子句,且子句會被解析並忽略。
樣本 – 清空分區
下列樣本從表sales的一個分區中刪除了資料。使用下列命令來建立表sales:
CREATE TABLE sales
(
dept_no number,
part_no varchar2,
country varchar2(20),
date date,
amount number
)
PARTITION BY LIST(country)
(
PARTITION europe VALUES('FRANCE', 'ITALY'),
PARTITION asia VALUES('INDIA', 'PAKISTAN'),
PARTITION americas VALUES('US', 'CANADA')
);
使用下列命令填充表sales:
INSERT INTO sales VALUES
(10, '4519b', 'FRANCE', '17-Jan-2012', '45000'),
(20, '3788a', 'INDIA', '01-Mar-2012', '75000'),
(40, '9519b', 'US', '12-Apr-2012', '145000'),
(20, '3788a', 'PAKISTAN', '04-Jun-2012', '37500'),
(40, '4577b', 'US', '11-Nov-2012', '25000'),
(30, '7588b', 'CANADA', '14-Dec-2012', '50000'),
(30, '9519b', 'CANADA', '01-Feb-2012', '75000'),
(30, '4519b', 'CANADA', '08-Apr-2012', '120000'),
(40, '3788a', 'US', '12-May-2012', '4950'),
(10, '9519b', 'ITALY', '07-Jul-2012', '15000'),
(10, '9519a', 'FRANCE', '18-Aug-2012', '650000'),
(10, '9519b', 'FRANCE', '18-Aug-2012', '650000'),
(20, '3788b', 'INDIA', '21-Sept-2012', '5090'),
(40, '4788a', 'US', '23-Sept-2012', '4950'),
(40, '4788b', 'US', '09-Oct-2012', '15000'),
(20, '4519a', 'INDIA', '18-Oct-2012', '650000'),
(20, '4519b', 'INDIA', '2-Dec-2012', '5090');
查詢表sales顯示了分區已由資料填充:
acctg=# SELECT tableoid::regclass, * FROM sales;
tableoid | dept_no | part_no | country | date | amount
----------------+---------+---------+----------+--------------------+--------
sales_europe | 10 | 4519b | FRANCE | 17-JAN-12 00:00:00 | 45000
sales_europe | 10 | 9519b | ITALY | 07-JUL-12 00:00:00 | 15000
sales_europe | 10 | 9519a | FRANCE | 18-AUG-12 00:00:00 | 650000
sales_europe | 10 | 9519b | FRANCE | 18-AUG-12 00:00:00 | 650000
sales_asia | 20 | 3788a | INDIA | 01-MAR-12 00:00:00 | 75000
sales_asia | 20 | 3788a | PAKISTAN | 04-JUN-12 00:00:00 | 37500
sales_asia | 20 | 3788b | INDIA | 21-SEP-12 00:00:00 | 5090
sales_asia | 20 | 4519a | INDIA | 18-OCT-12 00:00:00 | 650000
sales_asia | 20 | 4519b | INDIA | 02-DEC-12 00:00:00 | 5090
sales_americas | 40 | 9519b | US | 12-APR-12 00:00:00 | 145000
sales_americas | 40 | 4577b | US | 11-NOV-12 00:00:00 | 25000
sales_americas | 30 | 7588b | CANADA | 14-DEC-12 00:00:00 | 50000
sales_americas | 30 | 9519b | CANADA | 01-FEB-12 00:00:00 | 75000
sales_americas | 30 | 4519b | CANADA | 08-APR-12 00:00:00 | 120000
sales_americas | 40 | 3788a | US | 12-MAY-12 00:00:00 | 4950
sales_americas | 40 | 4788a | US | 23-SEP-12 00:00:00 | 4950
sales_americas | 40 | 4788b | US | 09-OCT-12 00:00:00 | 15000
(17 rows)
要刪除分區americas的內容,先要調用下列命令:
ALTER TABLE sales TRUNCATE PARTITION americas;
現在,查詢表sales顯示了分區americas的內容已被刪除:
acctg=# SELECT tableoid::regclass, * FROM sales;
tableoid | dept_no | part_no | country | date | amount
--------------+---------+---------+----------+--------------------+--------
sales_europe | 10 | 4519b | FRANCE | 17-JAN-12 00:00:00 | 45000
sales_europe | 10 | 9519b | ITALY | 07-JUL-12 00:00:00 | 15000
sales_europe | 10 | 9519a | FRANCE | 18-AUG-12 00:00:00 | 650000
sales_europe | 10 | 9519b | FRANCE | 18-AUG-12 00:00:00 | 650000
sales_asia | 20 | 3788a | INDIA | 01-MAR-12 00:00:00 | 75000
sales_asia | 20 | 3788a | PAKISTAN | 04-JUN-12 00:00:00 | 37500
sales_asia | 20 | 3788b | INDIA | 21-SEP-12 00:00:00 | 5090
sales_asia | 20 | 4519a | INDIA | 18-OCT-12 00:00:00 | 650000
sales_asia | 20 | 4519b | INDIA | 02-DEC-12 00:00:00 | 5090
(9 rows)
儘管刪除了記錄,分區americas的結構仍然是完整的:
acctg=# SELECT partition_name, high_value FROM ALL_TAB_PARTITIONS;
partition_name | high_value
----------------+---------------------
europe | 'FRANCE', 'ITALY'
asia | 'INDIA', 'PAKISTAN'
americas | 'US', 'CANADA'
(3 rows)