Identity columns

If you are inserting data from some other source to a table with an identity column and you need to ensure you retain the indentity values, you can temporarily allow inserts to the indentity column. Without doing so explicitly you will receive an error if you attempt to insert a value into the indentity column.

For example, if I have a table named EMP and I want to allow inserts into it's identity column, I can execute the following:


set identity_insert EMP on

Once you execute the command you will be able to insert values into the table's identity column. This will stay in effect in until you turn it off by executing the following:


set identity_insert EMP off

Be aware that at any time, only a single table in a session can have the identity_insert set to on. If you attempt to enable this for a table and another table already has this enabled, you will receive an error and will not be able to do so until you first turn this off for the other table. Also, if the value used for the indentity is larger than the current identity value then the new value will be used for the identity seed for the column.


Reseeding the identity value:

You can reseed the indentity value, that is, to have the identity values reset or start at a new predefined value by using DBCC CHECKIDENT. For example, if I have a table named MYTABLE and I want to reseed the indentity column to 30 I would execute the following:


dbcc checkident (EMP, reseed, 30)

If you wanted to reseed the table to start with an identity of 1 with the next insert then you would reseed the table's identity to 0. The identity seed is what the value is currently at, meaning that the next value will increment the seed and use that. However, one thing to keep in mind is that if you set the identity seed below values that you currently have in the table, that you will violate the indentity column's uniqueness constraint as soon as the values start to overlap. The identity value will not just “skip” values that already exist in the table.