Objectives of the new Infor BI OLAP Server Rules Engine

Accelerators were introduced early in the development of the OLAP Server, because rule calculations were dramatically slower than aggregations. Based on the accelerators, the OLAP Server generates, at cube load time or while adding a value to a cell, one or more flags to mark those cells that need to be calculated at runtime. Under the old OLAP Server Rules Engine, the administrator of the model is responsible for writing accelerator definitions to flag the cells to be calculated. This is a complex and error prone process.

The goal of the new OLAP Server Rules Engine is to speed up the non-accelerated rule calculations in such a way that the performance is comparable to the performance of an accelerated cube. As a consequence, the creation of rules is much easier and the error prone process of creating accelerators is no longer required.

Additionally, omitting the accelerators avoids the requirement to create accelerator flags at load time. This speeds up the cube loading dramatically.

The new OLAP Server Rules Engine works with the same rules syntax as the old OLAP Server Rules Engine. The intention is to make it easier for users to switch to it.

For compatibility reasons existing accelerated rules continue to work, meaning that customers are not forced to change their rules. Nevertheless the performance without accelerators can be better than with accelerators, i.e. please try to use your rules with and without accelerators to optimize your implementation.