Advanced Cube Design - Professional SQL Server Analysis Services 2005 with MDX - page 478

Summary
Ok, Neo, crawl up out of the rabbit hole; you have reached the conclusion of
this chapter. And what a profound trip that was! This chapter has provided you
with amazing new tools for your BI repertoire. In fact, consider yourself admit-
ted to the knowledgeable inner-circle of BI professionals because you know
that you can add measure groups from any accessible cube to your own cube
without the use of views in SQL Server. This is functionally equivalent to the
virtual cube scenario in Analysis Services 2000 without the extra steps associ-
ated with creation and maintenance of all those sub-cubes. This one change
encapsulates a lot of power, and you are going to have fun discovering that
power. Similarly, with remote partitions there is that compelling scalability
factor that kicks in; it's great! The Cube Wizard also provides you the ability to
create Cubes without a data source through templates similar to the Dimen-
sion Wizard which was not discussed in this chapter. The Cube Wizard has
two templates to choose from. Once the Cube has been created you would
need to populate the tables with appropriate data before processing the cube.
We leave it to you to explore this option from the Cube Wizard by selecting the
build method "build a cube without a data source."
Much of this chapter was dedicated to cube enhancements, which fall directly
to the bottom line of providing business information; using actions and KPIs
both will enhance any digital dashboard you might create. In fact, when you
start building custom front ends for consumption of your business intelligence
applications, like digital dashboards which reflect current business operations,
there are cases where you will have to write your own application for filtering
based on real-time data. Indeed, some complex and custom operations cannot
even be defined in a cube. For such operations, Analysis Services provides
support for the writing of custom code in ActiveX components or in managed
code. There is more on the programmatic approach in the next chapter .