Deleting Duplicate Rows

To delete duplicate rows in Sql server is one of the tedious tasks. Duplicate rows might exist in a table because of bad database design or because constraints are not applied. Unlike Oracle where we get unique row-id for each row and using that we can delete duplicate records, there is no simple way to delete duplicate records in Sql Server.

One method used to delete duplicate records is to write a stored procedure, use cursor in that and delete row one by one. There is another simple interactive way to delete duplicate rows in Sql Server without using stored procedures and cursor.

I will use Enterprise manager for explanation but scripts can be used if duplicate records are to be deleted on regular basis from some table. Create one table OrderDetail with columns ShipmentId, OrderId, ArticleId and Quantity where ShipmentId, OrderId and ArticleId fields should be unique.

Add few duplicate records into this table.

Create a blank copy of table using:

Select * into OrderDetailCopy from OrderDetail where 1 = 0

Create Unique index on columns ShipmentId, OrderId and ArticleId. Also check Ignore duplicate key checkbox. Save the table.

insert into OrderDetailCopy Select * from OrderDetail

You will get warning message:

Server: Msg 3604, Level 16, State 1, Line 1
Duplicate key was ignored.

Now you have OrderDetailCopy table without any duplicate rows.

Drop table OrderDetail using

Drop table OrderDetail

Rename table OrderDetailCopy to OrderDetail using

Sp_rename 'OrderDetailCopy','OrderDetail'