A target load order constraint restricts how the Data Integration Service loads and commits rows to two target instances related to each other in the same mapping.
In the Developer tool you can configure constraints to restrict the order that the Data Integration Service loads rows to target tables.
You can configure a constraint to force the Data Integration Service to load the data of a primary target instance completely before loading data to a secondary target instance. The tables that you define as the primary target and secondary target depend on the transactions in the input rows.
Consider the following scenarios for target load order constraints:
Insert rows to a master and a detail target.
You might configure a target load order constraint when you are inserting rows to targets that have a primary key-foreign key relationship. Configure the target with the primary key as the primary target instance. Configure the target with the foreign key as the secondary target instance. The Data Integration Service can stage the data for the secondary target until it completes loading the primary target.
Delete rows from a master and a detail target.
When you need to delete rows from targets with a primary key-foreign key relationship, you configure a different constraint. Configure the target with the foreign key as the primary target instance to delete the rows from the detail target first. Configure the target with the primary key as the secondary target instance.
Insert rows and update rows to the same relational table.
You can configure a target load order constraint for a mapping that loads insert rows and update rows to a relational table from two separate transformations. Configure the constraint to restrict the Data Integration Service from loading the update rows until after it loads the insert rows.
Target load order for flat files.
You can configure a target load order constraint for a mapping that loads rows into multiple flat file targets. Configure the target load order to load the secondary flat file after the primary flat file.
You can configure multiple constraints in a mapping. The Data Integration Service determines the most efficient execution plan to load the targets without violating the constraints.