If we created a metadata table that contained each possible dimension, it would be easy to write a PL/SQL routine to dynamically re-create and populate the 45 summary tables that are required. The pseudo-code for such a routine might look something like this:

DECLARE CURSOR c1 AS
    SELECT dimension_name FROM dimensions ORDER BY dimension_name;
FOR EACH outer_dimension
REPEAT
   OPEN CURSOR c1;
   FETCH c1 INTO dimension_one;
   DECLARE CURSOR c2 AS
   SELECT dimension_name FROM dimensions ORDER BY dimension_name;
   FOR EACH inner_dimension
   REPEAT;
   OPEN CURSOR c2;
   FETCH c2 INTO dimension_two;
   CALL create_SQL (dimension_one, dimension_two);
   NEXT inner_dimension
NEXT outer_dimension

In this fashion, we could create all 45 summary tables each night immediately after our new values have been loaded. But will all of these tables be needed by our end users? Because a complete pre-calculation of 45 tables is not an overwhelming task, we may decide that we can afford to have all possible summary table combinations available to our end users.

On the surface, it appears that SQL can be used against two-dimensional tables to handle three-dimensional time-series problems. It also appears that SQL can be used to roll up aggregations at runtime, alleviating the need to do a roll up at load time, as with a traditional database. While this implementation does not require any special multidimensional databases, the following two important issues remain to be resolved:

* Performance--Joining a table against itself (especially when comparing ranges of dates) may create many levels of nesting in the SQL optimization resulting in poor response time.

* Ability--No end user would be capable of formulating this type of sophisticated SQL query.

If you strip away all of the marketing hype and industry jargon, you can see that a data warehouse and a multidimensional database can be easily simulated by pre-creating many redundant tables, each with pre-calculated roll-up information. In fact, the base issue is clear--complex aggregation needs to be computed at runtime or when data is loaded.