Validating Your Data Model

Inconsistent, incorrect, or missing metadata can lead to problems when processing data or creating reports. Running the model scan validates your data model and alerts you to any issues. After creating a data model, a Space Administrator can run a model scan from the command line using the validatemetadata command.

Example metadata validation rules include:

Hierarchy names should be unique and level names should be unique within the hierarchy.

The expression for a Custom Attribute or Custom Measure cannot exceed 10 nested IF/ELSE statements as this creates more than 10 levels of nested case statements which is not supported by SQL Server.

To validate your data model against all rules, use the following syntax: validatemetadata all

The model validation scans the repository against a set of validation rules and provides a detailed report that describes any issues that are detected. You can run the scan multiple times against your data model. You can select the areas to check such as data sources, hierarchies, aggregates, and custom attributes or select specific rules to validate against your data model.