This document outlines the steps required to create a schema-only backup of a database. The backup can then be used to baseline a project, see Backups as baselines for more information.

1 - Create a schema only copy of the database

The DBCC CLONEDATABASE command creates a schema-only, read-only copy of a database. The example below creates a schema-only copy of the AdventureWorks2022 database called AdventureWorks2022_Copy:

DBCC CLONEDATABASE(AdventureWorks2022, AdventureWorks2022_Copy) with VERIFY_CLONEDB;

The VERIFY_CLONEDB argument verifies the consistency of the new AdventureWorks2022_Copy database. Please see here for more information about the limitations of DBCC CLONEDATABASE command and the list of objects it supports.

2 - Create a backup of the copied database

Once a schema-only copy of the database exists, then a backup of that database can be created. The example below creates a backup of the AdventureWorks2022_Copy database and saves it in the /var  folder:

BACKUP DATABASE [AdventureWorks2022_Copy] TO DISK = N'/var/AdventureWorks2022_Copy.bak';

3 - Drop the copied database

Once the backup exists, the copied database is no longer needed and can be dropped:

DROP DATABASE [AdventureWorks2022_Copy];