INSERT ON DUPLICATE KEY UPDATE can't update same row twice - mysql

Question

I'm trying to use sql in Mysql
INSERT INTO product_sales (product_code,product_desc,product_quantity,product_sales,product_group,insert_time)
VALUES ('ZSHA','AAA','1','1.55','TESTING','$TEST time')
ON DUPLICATE KEY UPDATE
product_quantity = VALUES(product_quantity) + '123',
product_sales = VALUES(product_sales) + '1.5',
product_desc = 'hello',
insert_time = 'hello';
Here is my Create table statement:
CREATE TABLE `product_sales` (
`product_code` varchar(20) NOT NULL DEFAULT '',
`product_desc` longtext,
`product_quantity` bigint(20) DEFAULT NULL,
`product_sales` float DEFAULT NULL,
`product_group` varchar(20) NOT NULL DEFAULT '',
`insert_time` varchar(40) DEFAULT NULL,
PRIMARY KEY (`product_code`,`product_group`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1
Here is the description of the table
| Field | Type | Null | Key | Default | Extra |
+------------------+-------------+------+-----+---------+-------+
| product_code | varchar(20) | NO | PRI | | |
| product_desc | longtext | YES | | NULL | |
| product_quantity | bigint(20) | YES | | NULL | |
| product_sales | float | YES | | NULL | |
| product_group | varchar(20) | NO | PRI | | |
| insert_time | varchar(40) | YES | | NULL |
Here is the problem, after i inserted the row, this query only update the data once in the database which means after the very first time update, the data won't update anymore?
What is the problem of this sql?

Answer 1

Seems to be working fine. There doesn't seem to be any problem with the SQL; it seems to be performing the specified operations; the behavior we observe is exactly as described in the documentation MySQL Reference Manual.
On the third execution, there are no changes to be applied; the resulting row will match exactly the row that is already stored. So MySQL reports 0 row(s) affected, and the row remains unchanged.
After first execution:
product_code product_desc product_quantity product_sales product_group insert_time
------------ ------------ ---------------- ------------- ------------- ----------
ZSHA AAA 1 1.55 TESTING $TEST time
After second execution:
------------ ------------ ---------------- ------------- ------------- ----------
ZSHA hello 124 3.05 TESTING hello
After third execution (and fourth, fifth, ...):
------------ ------------ ---------------- ------------- ------------- ----------
ZSHA hello 124 3.05 TESTING hello
These results are exactly as we would expect. It's not at all clear why would expect something different. (Maybe we're not understanding what the special VALUES() function does?)
The question leaves us guessing... what behavior is being observed, and what behavior is expected. Saying that it "is not working" is next to useless, in terms of explanation.
If we want to add the value supplied in the VALUES clause to the existing value in the column...
INSERT INTO ... ( ... , product_quantity , ... )
VALUES ( ... , '1' , ... )
ON DUPLICATE KEY
UPDATE product_quantity = IFNULL(product_quantity,0) + VALUES(product_quantity)
When a duplicate key exception occurs, this will get the current value of the product_quantity column from the row. If it's null, we return 0 otherwise we return the value stored in the row. Then add to that the value supplied in the INSERT statement for product_quantity.
If we don't want a NULL value to overwrite the current value in the column, we can wrap that in an IFNULL or COALESCE as well
...
ON DUPLICATE KEY
UPDATE product_quantity = IFNULL(product_quantity,0) + IFNULL(VALUES(product_quantity),0)
We're guessing at the specification; guessing at what the SQL is supposed to do, what behavior we want to achieve. Without that, we're just flinging "try this" SQL without anyway of determining if the proposed SQL satisfies the specification.

Answer 2

This is the way you have to do your insert statement if you want to keep adding values to columns product_quantity and product_sales, first remove VALUES() method, and second (maybe optional) for skip auto cast of types on engines, use integers and float numbers on the SUM.
INSERT INTO product_sales (product_code,product_desc,product_quantity,product_sales,product_group,insert_time)
VALUES ('ZSHA','AAA','1','1.55','TESTING','$TEST time')
ON DUPLICATE KEY UPDATE
product_quantity = product_quantity + 123,
product_sales = product_sales + 1.5,
product_desc = 'hello',
insert_time = 'hello';
I let you a description of the meaning of VALUES() method next, since i think you misunderstand it:
VALUES() Explained:
In an INSERT ... ON DUPLICATE KEY UPDATE statement, you can use the VALUES(col_name) function in the UPDATE clause to refer to column values from the INSERT portion of the statement. In other words, VALUES(col_name) in the UPDATE clause refers to the value of col_name that would be inserted, had no duplicate-key conflict occurred. This function is especially useful in multiple-row inserts.