Skip to main content

What is transaction in SQL Server

Transaction is the collection of T-SQL statements which executes in single unit. Transaction begins with a specific T-SQL statement and ends when all T-SQL statements execute successfully. If any statements fail then transaction fails. Transaction has only two types of results success or failed.

If transactions statements execute successfully then transaction is committed and changes saved successfully in database. If any statement failed then complete transaction become failed and rollback all statements execution.

When we use Transaction in SQL Server

If more than one table/view are related to each other and if we make any changes on any table/view and this change should be on related table also, here we have to use transaction.

Use transaction carefully and use it when needed because when transaction begins it locks all tables which are used in T-SQL Statement for data isolation. During transaction execution period no one can perform any operation on used tables in transaction due to region of maintaining data integrity.
Transaction use the following statements
  •          Begin transaction
  •          Commit Transaction
  •          Rollback Transaction

Example:

Suppose we have two table Users and UserProfile as bellow. Execute the following script


CREATE TABLE Users
(
ID BIGINT PRIMARY KEY,
NAME VARCHAR(100),
EMAIL VARCHAR(100),
PWD VARCHAR(100),
CREATED_DATE DATETIME
)

CREATE TABLE UserProfile
(
            ID BIGINT PRIMARY KEY,
            USERID BIGINT foreign Key references Users(ID),
            STREET VARCHAR(100),
            CITY VARCHAR(100),
            STATE VARCHAR(100),
            COUNTRY VARCHAR(100),
            PHONE VARCHAR(100),
            CAREATED_DATE DATETIME,
            MODIFIED_DATE DATETIME
)

INSERT INTO [dbo].[Users](ID,NAME,EMAIL,PWD,CREATED_DATE)
VALUES(1,'Codefari','info@codefari.com','codefari@123',GETDATE())                     

INSERT INTO [dbo].[UserProfile](ID,USERID,STREET,CITY,STATE,COUNTRY,PHONE,CAREATED_DATE,MODIFIED_DATE)
VALUES(1,1,'G-234','EAST DELHI','DELHI','INDIA','9089987767',GETDATE(),GETDATE())

SELECT * FROM Users
SELECT * FROM UserProfile



Result: 









Now we will try to insert data in both tables using transaction.
Users table have a column ID which is primary key and it is reference key in UserProfle table. UserProfile table ID is also primary key. In following script I will put existing UserProfile .ID value and it will generate an error and transaction should be failed and return same as above result. It means no any changes in Users table also.


BEGIN
BEGIN TRANSACTION trn
            BEGIN TRY
            --Insert data in Users Table
                        INSERT INTO [dbo].[Users](ID,NAME,EMAIL,PWD,CREATED_DATE)
                        VALUES(2,'Dilip','dilip@gmail.com','dilip@123',GETDATE())      
                       
            --Insert data in UserProfile table where ID is already exist                  
                        INSERT INTO [dbo].[UserProfile](ID,USERID,STREET,CITY,
                        STATE,COUNTRY,PHONE,CAREATED_DATE,MODIFIED_DATE)
                        VALUES(1,2,'G-234','EAST DELHI','DELHI','INDIA',
                        '9089987767',GETDATE(),GETDATE())
                        COMMIT TRANSACTION
            END TRY
            BEGIN CATCH
                                    ROLLBACK TRANSACTION
            END CATCH      
END

SELECT * FROM Users
SELECT * FROM UserProfile


Result:








In above script I put the UserDetails.Id value 1 so it’s generate an error and rollback transaction now in following script I changed the unique value with 2 and execute again.


BEGIN
BEGIN TRANSACTION trn
            BEGIN TRY
            --Insert data in Users Table
                        INSERT INTO [dbo].[Users](ID,NAME,EMAIL,PWD,CREATED_DATE)
                        VALUES(2,'Dilip','dilip@gmail.com','dilip@123',GETDATE())      
                       
            --Insert data in UserProfile table where ID is already exist                  
                        INSERT INTO [dbo].[UserProfile](ID,USERID,STREET,CITY,
                        STATE,COUNTRY,PHONE,CAREATED_DATE,MODIFIED_DATE)
                        VALUES(2,2,'G-234','EAST DELHI','DELHI','INDIA',
                        '9089987767',GETDATE(),GETDATE())
                        COMMIT TRANSACTION
            END TRY
            BEGIN CATCH
                                    ROLLBACK TRANSACTION
            END CATCH      
END

SELECT * FROM Users
SELECT * FROM UserProfile


Result:










Everything fine transaction commited.

Related Link

Nested Transaction in SQL Server




Popular posts from this blog

Add day to ISODate in MongoDB

We can use $add operator to add days in ISODate in mongodb, $add is the Arithmetic Aggregation Operator which adds number and date in mongodb.
Syntax:

{ $add: [ <expression1>, <expression2>, ... ] }

Note:  If one of the argument is date $add operator treats to other arguments as milliseconds to add to the date.
Example: Suppose we have a Test collection as below.

{"Title" : "Add day to ISODate in MongoBD","CreatedDate" : ISODate("2016-07-07T08:00:00.000Z")}

Query to add 2 days in CreatedDate

db.Test.aggregate([      { $project: { Title: 1, AddedDate: { $add: [ "$CreatedDate", 2*24*60*60000 ] } } }    ])

Result:

{ "_id" : ObjectId("579a1567ac1b3f3732483de0"), "Title" : "Add day to ISODate in MongoBD", "AddedDate" : ISODate("2016-07-09T08:00:00.000Z") }

Note: As mentioned in above note we have to convert days in millisecond because $add operator treat to other arg…

What is difference between UNION and UNION ALL in SQL Server

We use UNION and UNION ALL operator to combine multiple results set into one result set.
UNION operator is used to combining multiple results set into one result set but removes any duplicate rows. Basically, UNION is used to performing a DISTINCT operation across all columns in the result set. UNION operator has the extra overhead of removing duplicate rows and sorting result.
UNION ALL operator use to combine multiple results set into one result set but it does not remove any duplicate result. Actually, this does not remove duplicate rows so it is faster than the UNION operator. If you want to combine multiple results and without duplicate records then use UNION otherwise UNION ALL is better.
Following some rules for using UNION/UNION ALL operator
1.The number of the column should be the same in the query's when you want to combine them. 2.The column should be of the same data type. 3.ORDER BY clause can be applied to the overall result set not within each result set.
4.Column name of …

Remove special character from string in MongoDB

Problem: Suppose wehave a collection and one field is type string contains some special character (like !@#$%) and we don’t want these special character.
Solution: We can easily remove the special character from field using script “replace(/[^a-zA-Z 0-9 ]/g, '')” in our query.  How can we remove special character from string using this script please see following example.
Example: Suppose we have a collection “EduSurvey “where we are collecting information from institutions.

{Name:"JB institute”, About:"This is good one collage for MBA", Information:"This $%%institute ##has good faculty etc$$"}
{Name:"MK institute”, About:"This is good one collage for MCA", Information:"This$$%# is the dummy text12"}
{Name:"MG institute”, About:"This is good one collage for B,Tech", Information:"This# institute@ has&* good infrastructure"}

Did you notice Information fields contains some special character so we…