Data virtualization

In data virtualization (http://en.wikipedia.org/wiki/Data_virtualization), the logical application data model is decoupled from the physical integration data models (the data sources). The data page serves as the singular reference point throughout the application, joining application data models and integration data models.

If you change, add, or remove an integration point from which a data page is sourced, you only have to modify, add, or remove a mapping data transform.

In the following example, the data page is of one type and the data source is of another, incompatible type.

You can identify a response data transform of the same type as the data page to map the data from the data source to the data page, making it usable for the application:

The data transform acts automatically on each reference to the data page using that data source and maps the data in the manner that you specify:

A data page can have multiple data sources that it uses in different circumstances, depending on the parameters it receives with each reference. See Manage multiple data sources for a data page.

Regardless of which data source your use case requires, the data page maps the data it receives to the one common application data model.