Mastering Data Warehouse Aggregates: Solutions for Star Schema Performance

4.2. Requirements for the Aggregate Navigator

Evaluating schema design, table size, and availability can be confusing for an end user. And since these evaluations should be made each time a query is run, how is a query to be saved in a report? The aggregate navigator solves these problems.

The aggregate navigator is an important part of the data warehouse architecture. It maintains its own inventory of aggregates, their size, and availability, and rewrites queries as appropriate. In doing so, it shields users from the table selection process and allows all queries and reports to be expressed in terms of the base schema.

Commercial products handle aggregate navigation in a variety of ways. Before you study the different styles of aggregate navigation, it is useful to establish a set of requirements for the aggregate navigation function.

4.2.1. Why an Aggregate Navigator?

Without an aggregate navigator, aggregates damage the understandability of the schema, threaten its acceptance by end users, weaken its efficiency, and introduce maintenance issues. If aggregates are queried at a time when they are not in synch with the base tables, they may also provide wrong results.

As you have seen, the appropriate table to respond to a given query is a function of (1) the dimensionality of the query, (2) the dimensionality of the aggregate schema design, (3) the relative size of the potential responders, and (4) their availability status. Weighing these factors can be a tedious and counterintuitive ...