SQL Server Rider

Ignore arithmetic overflow error – SET Statement – SQL Server

Ayyappan Thangaraj — Sat, 18 May 2013 04:47:43 +0000

ARITHIGNORE Set statement is used to control the messages raised by arithmetic overflow or divide-by-zero errors during a query execution.

Usage

SET ARITHIGNORE ON

SET ARITHIGNORE OFF

SET ARITHIGNORE setting only controls whether an error message is returned or not. IF set ON then no message else if set OFF message will be returned. NULL value is returned in both case.

The SET ARITHABORT setting is used to determine whether the query is terminated in the event of arithmetic overflow or divide-by-zero error.

If SET ARITHABORT or SET ARITHIGNORE is OFF and SET ANSI_WARNINGS is ON then SQL Server still returns an error message when encountering divide-by-zero or overflow errors.

These set statement values are initialized by the SQL server while we create the session. So, It is mandatory to set all set statements properly to obtain the required result.

Get the current value of the set statements in the current session using @@OPTIONS.

Example

1. When ARITHABORT and SET ANSI_WARNINGS is OFF, ARITHIGNORE is ON. Null is returned due to the arithmetic errors.

2. When ARITHABORT and SET ANSI_WARNINGS is OFF, ARITHIGNORE is OFF. Null is returned due to the arithmetic errors and error message is also returned.

Set auto rollback transactions due to arithmetic overflow – SET Statements – SQL Server

Ayyappan Thangaraj — Fri, 17 May 2013 06:38:13 +0000

SET ARITHABORT statement allows SQL Server to end the query execution whenever the system encounters arithmetic overflow or divide-by-zero error. This set statement is always on by default. But, we can turn this off. If it is off then we may face poor performance problem and even query optimizer get different query.

Arithbort set statement becomes more effective if we use it with ansi_warnings set statement.

SET ARITHABORT ON + SET ANSI_WARNINGS ON = the query execution will be terminated if the system encounters arthimetic overflow or divide-by-zero error.

SET ARITHABORT ON + SET ANSI_WARNINGS OFF = This will terminate SQL execution and If the errors occur in a transaction then the transaction will be rolled back.

SET ARITHABORT OFF + SET ANSI_WARNINGS OFF = this will display a warning message and NULL is assigned to the result of the arithmetic operation.

note: If SET ARITHABORT is OFF, CREATE, UPDATE, INSERT, and DELETE statements on tables with indexes on computed columns or indexed views will fail. So, ARITHABORT must be ON when we create or change indexes on computed columns or indexed views.

Control SQL locks using locking statements – SQL Server – SET Statements

Ayyappan Thangaraj — Wed, 15 May 2013 07:08:50 +0000

Locking is an unavoidable activities in concurrency system. But, we can control and manage the resource effectively. But, in this blog post I am going to introduce a SET statement called DEADLOCK_PRIORITY. This is one of the locking set statement in SQL Server. It tells how to react when dead lock occurs in the current session. As the name priority suggest that the importance or the order in which the deadlock should be treated with other processes.

DEADLOCK_PRIORITY Syntax (SQL Server 2012)

SET DEADLOCK_PRIORITY { LOW | NORMAL | HIGH | | @deadlock_var | @deadlock_intvar }

::= { -10 | -9 | -8 | … | 0 | … | 8 | 9 | 10 }

Argument description

LOW, NORMAL and HIGH are the constant value pre-defined in SQL Server. The numeric value equivalent to the names are -5, 0 and 5 respectively. We can also use this strings in a variable to assign to the SET statement. And, We can also give any whole number value between -10 to 10 instead variable or constant. Setting low value means that the priority for the deadlock for that session is low and make the current session as deadlock victim.

Normal is the default value for DEADLOCK_PRIORITY.

SET statement examples

SET DEADLOCK_PRIORITY HIGH

Declare @dlVar as varchar(10) = ‘LOW’

SET DEADLOCK_PRIORITY @dlVar

SET DEADLOCK_PRIORITY -1

Deadlock victim

To resolves the deadlock by choosing one of the sessions as a deadlock victim. The current transaction of the deadlock victim is rolled back and deadlock error message 1205 is returned to the client. It helps the other session to continue its work without interruption. Priority value plays the vital role in choosing deadlock victim session.

How does this SET statement react to the deadlock in concurrency system?

When Session A’s DEADLOCK_PRIORITY is set with a value (-10 to 10) and If Deadlock occurred and the other session in the deadlock chain is having the DEADLOCK_PRIORITY value greater than Session A’s priority value then the Session A will be the deadlock victim.

if both sessions having same priority value than the SQL Server will choose the deadlock victim by estimating the cost to rollback the current transaction and choose the less cost session as deadlock victim.

I have already wrote a blog post about lock timeout set statement. Please read Lock_Timeout in SQL Server

Change the Date format of the current language – SQL Server – SET Statements

Ayyappan Thangaraj — Tue, 14 May 2013 06:59:53 +0000

Formatting date is one of the mandatory task for the developers to show date values in the user reports. There are various ways to format date in SQL Server database and reporting services. But, in this blog post I am going to show you how to set up the date format that user is going to use for the date value with rearranging the date part string (dmy) in SQL Server.

This setting only applies to date data type (date, smalldatetime, datetime, datetime2 and datetimeoffset) string value in an expression.

DateFormat set statement also overrides the date format defined in the default language settings. Learn about Language in SQL Server

Set statement syntax

SET DATEFORMAT { format | @format_var }

Format or @format variable must contain day (d), month (m) and year(y) combination format. For instance mdy, dmy, ymd, ydm, myd, and dym.

There are some restrictions applies to specific DateFormat format to use a date value stored in date, datetime2 and datetimeoffset data types. Because, we cannot use ydm format to use the date value.

A valid expression

Control current session and specific information – SQL Server – SET Statements

Ayyappan Thangaraj — Mon, 13 May 2013 00:31:08 +0000

SET statement in SQL Server is used to handle a specific information (such as date format, displaying text length and etc) in the current session by altering the current session’s default set values. Some of the SET statements are already setup by the administrator at SQL instance level. So, when we initiate the session it applies all the default SET options to the user of that session. But, we have an option to alter that default value. Usually, we have to turn the set option ON or OFF and in some cases we have to pass some value to the set option. There are different groups of SET statements available in SQL Server as given below.

I have discussed some of the functions that are related to SET statements in my blog post. Those are given below for reference.

Construct SQL statements using catalog views – SQL Server

Ayyappan Thangaraj — Fri, 10 May 2013 15:07:43 +0000

I recently worked on a project where there are some SQL statement that I need to write to insert, update and Delete the data in all columns in a set of tables. There may be some efficient way to implement this task. But, I found one simple solution for my scenario.

I have used SQL catalog views to build my SQL Statements. My scenario was to update the columns by replacing a character in it.

Example for building SQL Queries using catalog view

select ‘UPDATE ‘ + OBJECT_NAME(object_id) +’ SET ‘ + name + ‘ = REPLACE(‘ + name +’, ”+”,””)’ from sys.columns
where left(OBJECT_NAME(object_id), 3) = ‘tbl’

This query generates SQL update statements to replace character + to nothing in each column in all the tables starts with ‘tbl’ that I specified in the where clause.

Execution

there could be some very good way to do this. but found it is very quick solution for me. This is not the stopping point we can also use other object catalog views to accomplish our task quickly and simply.

object catalog views for reference

A simple way to dowload a web page with infinite timeout value using SSIS – SQL Server

Ayyappan Thangaraj — Fri, 10 May 2013 05:04:25 +0000

In this blog post I am going to show you a simple way to download a web page. I decided to write this blog post for one reason. HTTP connection manager does not allow us to configure the timeout value as we need. Because, we may need to set up different timeout value for different web request. So, you can use this VB.net code to download a webpage with custom timeout value.

We can also use this code for Web scraping .

Code snippet

Download source code for web scraping

You can also refer to the other blog posts that I wrote about similar topic

extend-the-http-connection-timeout-value-ssis

download-a-web-page-using-ssis

Import variable record length CSV or any delimited file to a database using SSIS – SQL Server

Ayyappan Thangaraj — Wed, 08 May 2013 22:32:32 +0000

Delimited text files are the method to store user data in a file system in the past and even now. These files are easy to create and simple to transfer from one system to another system without trouble. In modern data technology we use various data base management system that keeps the data in tabular format and the latest Big data evolution keeps data in key and value pair or in No-SQL database. Now, we have got delimited files and database system to store. But, We also need to migrate data from the text file to database using data export utility. These utilities are very well designed to handle the delimited file with single data structure (ie fixed record length).

How do we import variable record length delimited file to a SQL database?

In this demo, I will be using VB.net code to read variable length CSV file data and store it in SQL tables.

Scenario:

I have a CSV file with ATM transactions of bank customers. This file contains two types of record with variable length. Customer information and ATM transactions are two types of record in this file. Now, I need to upload this file to tables tblCustomer and tblTransactions in SQL Server database.

We already know that we have to use script task to write .net code to read the file and split the line with delimited character and upload in to the tables respectively. I do construct the SQL insert statement in side the script task for data upload.

Sample CSV File

Download CSV file (Bank_AX.TXT)

Table structure

tblCustomer

CREATE TABLE [dbo].[tblCustomer](
[Rec_ID] [varchar](50) NULL,
[AC_No] [varchar](50) NULL,
[Ac_Name] [varchar](50) NULL,
[TransactionDate] [varchar](50) NULL
) ON [PRIMARY]

tblTransaction

CREATE TABLE [dbo].[tblTransaction](
[Rec_ID] [varchar](50) NULL,
[Ac_No] [varchar](50) NULL,
[Transaction_Type] [varchar](50) NULL,
[Amount] [varchar](50) NULL,
[Transaction_Time] [varchar](50) NULL
) ON [PRIMARY]

Script

This image shows partial script. You can download the full code below this image.

Full source code for reading variable length records is available for download here.

Output

Data has been uploaded to tblCustomer and tblTransaction tables after executing this script.

Check or modify the current identity value in a table using DBCC – SQL Identity – SQL Server

Ayyappan Thangaraj — Mon, 06 May 2013 21:58:57 +0000

DBCC CHECKIDENT (): This statement is used to check the current identity value of a table. This function also used to reset the current seed value or re-assign new seed value to an identity column.

Function Syntax

DBCC CHECKIDENT
(
table_name
[, { NORESEED | { RESEED [, new_reseed_value ] } } ]
)
[ WITH NO_INFOMSGS ]

Arguments and examples to use this command

We can use this function in ways using the function arguments. NoReSeed, ReSeed with new value and No_Infomsgs are the optional value that we can use with executing the command.

Table_name: This is the valid table name used in this command within single quotes.

eg: DBCC CHECKIDENT ( ‘Table_name’)

NoReSeed: It will return the current identity value and the latest maximum identity value of an identity column in a table. by default, these two values should be same. otherwise, use reseed to do so.

NoReSeed: This argument will reset the identity value if the identity value is less than the latest maximum identity value.

eg: DBCC CHECKIDENT ( ‘Table_name’, RESEED ) or DBCC CHECKIDENT ( ‘Table_name’)

ReSeed with new Seed value: This argument will reinitialize the seed value of an identity column in a table.

eg: DBCC CHECKIDENT ( table_name, RESEED, new_reseed_value )

WITH NO_INFOMSGS : This argument value help to Suppresses all informational messages from the command.

Example

In this demo , I am going to fill the identity value gap in a table.

Step 1: Create a Table for this demo

CREATE TABLE tblIdentity
(
RowID INT IDENTITY,
Code VARCHAR(20)
)
GO

Step 2: Insert 5 rows in to the table using batch execution

INSERT INTO tblIdentity VALUES (‘SQL Server Rider’)
GO 5

Step 3: List or select all value from the table

SELECT * FROM tblIdentity
GO

Step 4: Delete a row that has the identity value 3

DELETE FROM tblIdentity WHERE RowID = 3
GO

Step 5: Check the current identity value in the table

DBCC CHECKIDENT(‘tblIdentity’, NORESEED)
GO

Step 6: Now, reseed the identity value to fill the gap. In this demo, the gap value is 3.

DBCC CHECKIDENT(‘tblIdentity’, RESEED, 2)
GO

Step 7: Check the current identity values. Now, we can see two different values in the current identity value and the current identity column value.

DBCC CHECKIDENT(‘tblIdentity’, NORESEED)
GO

Step 8: Insert a row to fill the gap.

INSERT INTO tblIdentity VALUES (‘SQL Server Rider’)
GO

Step 9: Now, execute this command to reset the identity value to the actual or correct value.

DBCC CHECKIDENT(‘tblIdentity’, RESEED, 5)
GO

DBCC CHECKIDENT(‘tblIdentity’, RESEED)

Output

Query execution

Result

Query execution Messages

Insert value to an Identity column – SQL Identity – SQL Server

Ayyappan Thangaraj — Fri, 03 May 2013 06:00:33 +0000

We all know that we cannot insert a value to an identity column in a table using insert statement. Yes, it is true. But, there is a way that allows us to explicitly insert and not update a value in the identity column.

SET IDENTITY_INSERT

This is a set statement that allows user to inserted a value into the identity column. It allows us to do so when this set statement is on otherwise we cannot insert value to identity column.

SET IDENTITY_INSERT Syntax

SET IDENTITY_INSERT [ database_name . [ schema_name ] . ] table { ON | OFF }

In this statement database name and schema name is optional. But, we have to use a valid table name along with on or off switch in the set statement. User should have valid permission to perform this operation.

Example

I have already designed a table with identity column. In this first example, I am going to update an identity column to see what is the error message does SQL Server return to the user.

Updating Identity column

Now, inserting a value in to an identity column

Now, I am going to do explicit insert to an identity column using IDENTITY_INSERT set statement.

Current values in the table

Delete and insert a row with value to an identity column. RowID column is the identity column in this table.

Now, I have inserted a row with value to an identity column. But, using this statement we can also insert duplicate identity value to a table. SQL Server will not check for duplicate identity value using when we use SET Identity_Insert and Insert statement. So, we have to use Primary or unique key constraint to keep valid identity values in the table.