If a database does not exist at the point that it is required (for example, you attempt to upload a data file to a datastore that has not yet been created in the database instance), it is created on-the-fly; this can be slow for certain RDBMS types, and so we recommended that you create any required databases ahead of time, using SQL scripts.
dbfhadmin -script -type:datastore -provider:ss -name:<datastore-name> [-db|existdb:<database-name>] [-file:<script-name.sql>]
where <datastore-name> is the name of the datastore to create; <database-name> is the name of a new database (-db) or existing database (-existdb) that can house multiple datastores - if you omit either option, a new database called <datastore-name> will be created for the datastore; <script-name.sql> is the name of the script file to be created .
Once you have created the script files, use the dbfhadmin utility to run them, and create the required databases. Before you run the -createdb command, ensure that the sqlcmd command is on the PATH.
dbfhadmin -createdb -usedb:<instance> -provider:ss -type:datastore -file:<script-name.sql> -user:<user> -password:<pass>
where <instance> is the SQL server instance name, and <user> and <pass> are valid logon credentials for the database. If -user and -password are not specified, SQL Server SSPI authentication is used - see your SQL Server documentation for more details.
Now that the datastores have been created, you can configure the database configuration file to use them.