To select target schema options, the target field mapping must be automatic.

When you configure target schema options for objects that are created at runtime,

Data Integration

creates the target the first time you run the task. In subsequent task runs,

Data Integration

updates the target based on the schema change option that you select.

The schema change handling options available are based on the target connection. To see if a connector supports dynamic schema change handling, see the help for the appropriate connector.

The following table describes the schema handling options that you can select for each target type:

Schema handling option	Target type	Description
Keep Existing File Format	File	Data Integration fetches the most recent target schema at runtime and does not apply upstream schema changes to the target file.
Drop Current and Recreate	Database and file	For database targets, Data Integration drops the existing target table and creates a new target table with the schema from the upstream transformations on every run. For file targets, Data Integration updates the target schema to match the incoming schema on every task run.
Alter and Apply Changes	Database	Data Integration updates the target schema with additive changes to match the schema from the upstream transformations. It does not delete columns from the target.
Don't Apply DDL Changes	Database	Data Integration fetches the target schema at runtime and does not apply upstream schema changes to the target table.

Data Integration

does not pass field constraints to the target. For example, the source contains fields S1 and S2 configured with the NOT NULL constraint. The target contains fields T1 and T2 also configured with the NOT NULL constraint. You select the

Alter and Apply Changes

schema handling option. When you run the task, fields S1 and S2 are written to the target with no constraints.