Database Reference
In-Depth Information
Handling Transactional-Level
Data
In this chapter, we will analyze how to add detailed information about each transaction
in a fact table, such as invoice document and line number. We'll compare the use of
MOLAP
(
Multidimensional Online Analytical Processing
) and
ROLAP
(
Relational
Online Analytical Processing
) dimensions for this purpose, and we will use the
drillthrough feature to expose this data to the end user. We will also explain the reason
this approach is better than exposing a large dimension directly to the end user.
In the second part of this chapter, we will add a dimension to the sales cube that
describes the reasons for a sale. Since each sale can have multiple reasons associated
with it, we will make use of the many-to-many dimensions relationship feature of
Analysis Services, discussing its properties and possible performance issues. We will
also take a brief look at possible modeling patterns available using many-to-many
dimension relationships.
Details about transactional data
The goal of a multidimensional cube is to analyze aggregated data across several
dimensions. However, when there is some interesting data, the user might be
interested in drilling down to a lower level of detail. For example, when it comes to
sales analysis, it could be interesting to look at the individual invoices that caused a
particular high volume of sales in a single month. This is a very common request for
end users to make; in fact, the question is not if the users will need this, but when.
Search WWH ::
Custom Search