By default, in order to satisfy a query against an OLAP cube, the SAS OLAP Server performs subqueries to obtain the data for each axis independently. It then submits a final subquery to complete the query result set.

In some cases, the last subquery to the physical data is redundant and might unnecessarily increase the response time to return the result set when the result set could instead be retrieved from the existing subquery cache.

Situations where this could occur include the following:

• when there are only measures on one axis and a subquery has already been performed on the other axis
• if there is only one axis in the query and it has had its initial subquery performed

Consider the following example:

SELECT
   [Measures].[TOTAL_AMOUNT] on columns,
   NON EMPTY
     ORDER( [PRODUCT].[All PRODUCT].CHILDREN ,
           [Measures].[TOTAL_AMOUNT], DESC)
   ON rows
FROM [ProductsCube]

This query produces the following two subqueries:

select distinct CATEGORY, S_TOTAL_AMOUNT from OLAPSIO.CAT
select distinct CATEGORY, S_TOTAL_AMOUNT from OLAPSIO.CAT

In this case, the second subquery could be avoided since the data for the final subquery is already available in the subquery cache.

Click the Hot Fix tab in this note to access the hot fix for this issue. Once the hot fix is applied, the final subquery will be evaluated to see if the result set can be returned from the subquery cache.

Operating System and Release Information

A fix for this issue for Base SAS 9.21_M3 is available at: