With the entire BWA model defined we can now start to deploy it. With the knowledge of how the source tables look like, what columns are used or omitted, where new columns are required to hold the foreign key of an relationships, the tool can generate the target data model.
This target data model is then validated. Either validated by the rules of Index Designer only or, what should be commonly done, validating the BWA rules as well. So we select the "validate against BWA" option.
The validation will hopefully show a green "OK" at the end.
With the syntax and logic checks away we can start deploying that into DataServices. Remember, right now nothing was done. No objects created in BWA, nothing. All of that will be done by DataServices, so we need to generate some objects there.
The datastore for BWA with its tables and the create/delete job are generated always usually. With these we can connect to BWA, with the create job a script is executed against BWA to create the tables there or with the delete job to get rid of them. So next we would need to build a job with multiple dataflows, each reading the source table and loading the data into the BWA table. Actually, Index Designer knows where the data comes from, what fields are loaded into BWA etc. Hence the tool has the option to generate even that for us if selected in the dialog.
As we said we want both, the code for the datastore, the three jobs (create/delete and load) are generated, the load job contains three dataflows, one for each target table.
Finally we might want to save our project into the DataServices repository and exit Index Designer. Back in DataServices, next step would be to run the create-job to instantiate the model in BWA and then run the load job to copy all rows from the DWH into BWA.