Logical Model versus Physical Data Model

Relationships between objects are maintained as part of the data that is stored in the platform database and are defined in two places.

  • Through the logical data model, CloudConnect users can define these dimensional relationships between the elements in your data sets. The logical data model defines how the data is organized.
  • When the logical data model is loaded into a project, it is used as the basis for determining how the physical data model is created for storing the data that is to be used in the project. This process enhances scalability, security, and overall performance of the datamart.
  • The physical data model is how the data is actually stored in the database. This method of storage is designed to achieve the best performance. For example, data normalization is automatically performed in the physical data model.

    The physical data model is automatically generated within the GoodData Platform. GoodData users and developers do not need to interact directly with it.

For a graphical representation, see Project Structures.

Within the GoodData Platform, the logical data model is unpacked to build the database tables used to store data. For example, each attribute in your logical data model becomes a separate database table. Each attribute label is turned into a separate column in the attribute's database table. In this manner, the physical data model is automatically normalized during the schema creation process.