Database Reference
In-Depth Information
Figure 10-21.
The Author-AuthorState relationship
Configuring the relationship to be flexible or rigid helps SSAS improve processing and query performance.
SSAS has a processing algorithm that runs more efficiently if it doesn't have to look for relationships that may
have varied over time. By setting the relationship to rigid, SSAS can skip over this section of the algorithm. Query
performance is affected by formulating an execution plan that is more efficient with a rigid relationship as well.
In our experience, 80 percent of all the relationships we define are rigid. An example of a rigid relationship is
data that does not change, such as the cities of a particular state. It is unlikely that Seattle, which is in the state of
Washington this year, will somehow be in the state of Oregon next year.
A flexible relationship example is sales territories. A specific sales territory may be assigned to one employee
and have a specific geographic area this year, but next year it may be assigned to a different employee.
Keep in mind that regardless of whether you choose flexible or inflexible, the processing and query results in
your reports will still work. This discussion is simply about efficiency and the benefits of selecting the appropriate
relationship type where it is known. When in doubt, however, you can leave it flexible. Failure to do so will have a
negligible impact on your dimension unless it is very large.
To map the AuthorID attribute to the Author attribute, repeat the process (Figure
10-22
).
