ArcEngine exploits the declared referential integrity constraints in a star schema to rewrite the query in a more optimal form. Specifically, ArcEngine rewrites the query to perform all necessary aggregations before making the join. This reverses the initial order of join followed by aggregation, to aggregation followed by join instead.
ArcEngine exploits single table analytical views, which are incrementally refreshable, on fact tables to make the query run faster. It also flattens the resulting query using merge rewrite, and exploits other optimizations, like run-time filters.
To fully enable the Eager Aggregation feature, you must configure the dashboard and the datasets:
Arcadia Enterprise Eager Aggregation has the following limitations:
It supports only simple queries based on a simple star schema, which uses primary and foreign key connections to define joins.
Fact and Dimension tables must be regular tables, not logical views.
Both primary and foreign keys must be simple columns, not expressions or multi-column constructs.