With the concept of applications or solutions leveraging a more "modular" data model approach, an idea I would have is closer integration of multiple Board data models. A simple example would be an FPA Planning model with an HR planning module. In this example I could see two different data models:
- Data model which contains the core planning pieces including the final P&L report.
- Data model which contains the HR planning. This would have entities that would not be incorporated into the P&L report such as "Employee" which is specific to the HR Planning module. There would also be security that is specific to this data model.
Currently if I set up multiple data models and try to sync between them I have a few options:
- In the source data model I need to export a cube or cubes into flat files and load these into the target data model via ASCII data readers.
- Leverage Board APIs to create flat files which I then have to create ASCII data readers into the target data model.
What I'm asking for is a way to bypass the step of creating flat files and having direct integration between Board data models:
- Potentially a new type of data reader (Board Data Reader) which allows me to connect to a data model on the same server instance.
- This would allow me to connect to a cube on the other data model and allow me to map appropriately to a cube on the target data model.
I could of course create procedures around this like any other data reader, but the key improvement here is skipping the flat file export and ASCII data reader step.
I think with more modular designs leveraging multiple data models, closer integration will be very beneficial moving forward.
