I would typically use Management Studio to rename columns or tables in SQL Server. Afterall, it is an easy thing to do to open up the GUI make your changes and then save them back to the SQL Server.

Have you ever been in a situation where you wanted to make a number of such changes and then later on you needed to repeat those changes on another system?

I was in such a situation recently where due to some application changes, many columns had to be renamed, the application tested and then I had to repeat the procedure on production.

I did not want to have to open up every table in Management Studio and click the save button each time all over again!

I needed something that I could execute quickly, so I prepared a file of T-SQL commands to make my changes in one swift stroke. I achieved it all using sp_rename

T-SQL example – renaming columns and tables

So here we have an example for renaming a column:

EXEC sp_rename 'Table.[OldColumnName]', '[NewColumnName]', '[COLUMN]';

If we wanted to rename a table then we can use:

EXEC sp_rename '[OldTableName]', '[NewTableName]';

Note that the square brackets are not mandatory but are necessary when you have silly database designers who have put such things as hyphens or spaces in their field names. Makes me cringe just thinking about it! 😐

Permissions needed for sp_rename

You need ALTER permission if you want to rename tables or columns.

One final note

You can use sp_rename for renaming other objects besides tables. It can also be used for renaming indexes, constraints, types databases and statistics. More more information on this useful little proc, visit this link

If you want to remove data from a table, you can do this in a couple of ways using TRUNCATE TABLE or DELETE. A common approach is to use DELETE to remove all rows in the table but TRUNCATE TABLE offers an alternative and I will list what the differences between the two are.

DELETE

--Delete rows with filter
DELETE FROM SchemaName.Table WHERE...

--Delete all rows
DELETE SchemaName.Table

--Delete all rows
DELETE FROM SchemaName.Table

--Delete using JOINS
DELETE T1
FROM SchemaName.Table1 AS T1
JOIN SchemaName.Table2 AS T2 ON T1.Column1 = T2.Column1

--Delete using IN
DELETE FROM SchemaName.Table1
WHERE Column1 IN (SELECT Column1 FROM SchemaName.Table2)

--Delete using EXISTS
DELETE T1
FROM SchemaName.Table1 AS T1
WHERE EXISTS
(SELECT *
FROM SchemaName.Table2 AS T2
WHERE T1.Column1 = T2.Column1)

• Provides more control over what data is and can be deleted.
• Will not reseed the identity value on an identity column
• Can be used where the DELETE operation does not violate foreign key constraint
• Can be a slow operation if the table is large
• Locking can be an issue if the operation is removing many rows
• Will fire triggers
• Transactions can be rolled back as LSN (log sequence number) is maintained in the transaction log
• Data pages consumed may need to be reclaimed using a DBCC SHRINKDATABASE operation
• DELETE will typically consume more transaction log space because of extra logging

TRUNCATE TABLE

TRUNCATE TABLE Schema.TableName;

• WHERE clause is not optional – it will remove all data from your table
• Reseeds the identify value of an identity column
• Cannot be used on a table which is referenced by a foreign key constraint
• Can be considerably quicker than DELETE due to deallocation of data pages and minimal logging
• Cannot be used in replication or log shipping
• Will not fire triggers
• Cannot be used on tables which are part of an indexed view
• Who has permission: table owner, sysadmin, db_owner, db_ddladmin or any user with ALTER permission on the table

I wanted to expand on the differences in speed between the two operations.

When TRUNCATE TABLE is issued against a table, all the data pages are deallocated which are used to store the tables data. Compared with a DELETE, this is significantly faster as each DELETE operation is logged in the transaction log but only the deallocations are logged in the transaction log for a TRUNCATE operation.

The DELETE operation will consume database resource and locks and can therefore be considerably slower and have greater impact on your system in comparison to TRUNCATE TABLE. See this post on reducing overhead for large delete operations.

As each DELETE operation is logged inside of the transaction log, it is possible to rollback to a point in time but with TRUNCATE TABLE it is not possible to have this level of precision. In a rollback scenario, only a complete rollback could be performed.

As with a DELETE it is also possible to rollback a TRUNCATE if it has not yet been committed. You can do this by using transactions like in this very simple example:

BEGIN TRAN
TRUNCATE TABLE SchemaName.Table
--test success here
ROLLBACK TRAN
ELSE
COMMIT TRAN

Time to wrap this up..

So there are a number of differences between the two. Both operations have their pros and cons and provide you with a choice over how you want to remove your data.

New in SQL Server 2012 as part of exception handling in T-SQL is THROW

Exception handling in previous versions of SQL Server

THROW will raise an exception inside of the catch block and there are already ways to do this in previous versions of SQL Server but where this is an improvement over previous methods is that it simplifies writing the code to raise them.

So lets compare the old and the new.

Here is the T-SQL:

SQL Server 2005/SQL2008

BEGIN TRY
DECLARE @VarToTest INT
SET @VarToTest = 'C'
END TRY
BEGIN CATCH
DECLARE
@ErrorMessage nvarchar(4000)
, @ErrorSeverity int

SET @ErrorMessage = ERROR_MESSAGE()
SET @ErrorSeverity = ERROR_SEVERITY()

RAISERROR (@ErrorMessage, @ErrorSeverity, 1 )
END CATCH

SQL Server 2012

BEGIN TRY
DECLARE @VarToTest INT
SET @VarToTest = 'C'
END TRY
BEGIN CATCH
THROW
END CATCH

Msg 245, Level 16, State 1, Line 4 Conversion failed when converting the varchar value ‘C’ to data type int.

For more information see this link from books online – SQL Server 2012 THROW