For many data models, particularly those that were developed as part of pilot projects or by business users, a number of modifications can be implemented to reduce resource requirements or improve query performance. Therefore, prior to concluding that a certain amount of Premium capacity (or Analysis Services resources) is required, data models can be evaluated against a number of standard design practices and optimization techniques such as the following:
- Avoid duplicate or near-duplicate data models:
- Design and maintain a consolidated, standardized data model of fact and dimension tables.
- Remove tables and columns that aren't needed by the model:
- For import mode models, columns with the unique values (cardinality) ...