Database Reference
In-Depth Information
Chapter 5. Optimizing Dimension and
Cube Processing
In this chapter, we will cover:
• Understanding dimension processing options
• Learning about basic dimension processing
• Learning advanced dimension processing options
• Using out-of-line bindings for dimension processing
• Dealing with partition processing options
• Using SQL Server Integration Services to process Analysis Services objects
• Monitoring and tuning processing performance
Introduction
This chapter discusses various processing options available with dimensions and
cubes. As discussed in the previous chapters, all commands sent to Analysis Ser-
vices use XMLA language—the same holds true for processing commands. You will
start by learning the basic processing commands and work your way up to more com-
plex options.
If you use ROLAP storage for any of your SSAS objects, processing simply defines
metadata; ROLAP does not copy relational data into multidimensional format. You can
reference Chapter 3 , Creating Analysis Services Cubes , for more information about
various storage options. For the remainder of this chapter, we will discuss processing
for MOLAP objects.
Processing loads the data from the relational database tables into the Analysis Ser-
vices objects. The easiest option allows the processing of the entire Analysis Services
database, including all dimensions and all partitions. This option is suitable for tiny
sample databases or for a proof of concept type implementation that has a few mega-
bytes of data. However, for real-world applications, you must approach processing
carefully, as loading many gigabytes or even terabytes of data into the cube daily is
not a viable option.
Search WWH ::




Custom Search