我有以下用于在我的数据库中插入新记录的SQL代码
DECLARE @CustomerID INT
DECLARE @PropertyID INT
BEGIN TRAN T1
INSERT INTO c_customer (title, f_name, l_name, tel1, tel2, tel3, email, email2,
type, primary_contact, tel1type, tel2type, tel3type)
VALUES(@title, @fname, @lname, @tel1, @tel2, @tel3, @email, @email2,
'Owner', 1, @teltype1, @teltype2, @teltype3)
SET @CustomerID = SCOPE_IDENTITY()
BEGIN TRAN T2
INSERT INTO c_property (address1, address2, address3, post_code, city, county)
VALUES (@address1, @address2, @address3, @postcode, @city, @county)
SET @PropertyID = SCOPE_IDENTITY()
UPDATE c_property
SET invoice_flag = @PropertyID
WHERE c_property = @PropertyID
BEGIN TRAN T3
INSERT INTO c_customer_assignment
VALUES (@PropertyID, @CustomerID)
COMMIT TRAN T1
COMMIT TRAN T2
COMMIT TRAN T3
SELECT @CustomerID, @PropertyID
此代码按照我的意愿运行,确保使用c_customer_assignment
表正确链接添加的详细信息,但它确实看起来过于复杂,我想知道我是否采取了正确的方法来解决问题(不确定我是否甚至需要嵌套事务)。
我知道至少需要进行一次交易,因为我需要确保记录不会因为其他用户同时插入而导致不匹配。
我是否也需要审查事务隔离,或者这还不够吗?
答案 0 :(得分:2)
您只需要一笔交易:
DECLARE @CustomerID INT
DECLARE @PropertyID INT
BEGIN TRAN T1
INSERT INTO c_customer (title, f_name, l_name, tel1, tel2, tel3, email, email2, type, primary_contact, tel1type, tel2type, tel3type)
VALUES(@title, @fname, @lname, @tel1, @tel2, @tel3, @email, @email2, 'Owner', 1, @teltype1, @teltype2, @teltype3)
SET @CustomerID = SCOPE_IDENTITY()
INSERT INTO c_property (address1, address2, address3, post_code, city, county)
VALUES (@address1, @address2, @address3, @postcode, @city, @county)
SET @PropertyID = SCOPE_IDENTITY()
UPDATE c_property
SET invoice_flag = @PropertyID
WHERE c_property = @PropertyID
INSERT INTO c_customer_assignment
VALUES (@PropertyID, @CustomerID)
COMMIT TRAN T1
SELECT @CustomerID, @PropertyID
答案 1 :(得分:0)
Use try catch instead of some many trasactions.. BEGIN TRANSACTION; BEGIN TRY ---your dml operation should go here --- END TRY BEGIN CATCH ---CATCH ERROR HERE AND ROLLBACK OPERATIONS HERE IF @@TRANCOUNT > 0 ROLLBACK TRANSACTION; END CATCH; IF @@TRANCOUNT > 0 COMMIT TRANSACTION; GO