Database Reference
In-Depth Information
medication data for a hospital. If a doctor checks a report to find out what medication a patient has been given,
the time between entering the medication dispensing data into the database and the time that it shows on the
report must be as short as possible.
When the cube partition is designed as ROLAP, report queries sent to the cube are passed through the SSAS
server to the data warehouse in real time. In our example, the data warehouse is an OLTP database that records patient
medications; therefore, reports through SSAS will show medication data as soon as it is entered into the OLTP database.
With a ROLAP design, you may ask yourself, “Couldn't I get the same reports directly from the OLTP
database?” The answer is, “yes!” And, in fact, if you only occasionally need immediate reports, creating a ROLAP
cube can be considered redundant.
Still, there some advantages to ROLAP. It is true that you bypass all the performance gains normally
associated with the SSAS server, but you still gain the capacity to rename columns, create user-defined
hierarchies, and associate dimensions with cube measures.
In the BI world, ROLAP designs are rare, and the same is true of the hybrid approach, HOLAP. HOLAP
cubes are useful on occasions when reports deal with aggregate values and not individual leaf values. Examples
would be reports designed as dashboards with KPIs. These reports require only the higher level, pre-aggregated
values, and not the leaf level data (which is stored back in the OLTP database). Therefore, report performance is
improved without having to store a copy of the leaf-level data on the SSAS computer.
Partition Storage Settings
The Partition Storage Settings dialog window allows you to configure a partition's storage design (as shown in
Figure 12-17 ). MOLAP is the most common choice, and it is divided into four categories. Microsoft has included
Figure 12-17. The Partition Storage Settings dialog window
 
Search WWH ::




Custom Search