InfoProvider Compression: What It Is and When to Do It
- by Gary Nolan, SAP BI Strategic Architect, Sapiex Associates LLC
- June 1, 2005
Do you compress data packages loaded into BW InfoProviders? SAP says you should, but why? Learn why the periodic compression of BW InfoCube data is a good idea.
InfoProviders have two types of fact tables to hold key figure data. Although both are similar, the E fact table stores compressed data, while the F fact table stores uncompressed data.
SAP recommends that you compress data packages loaded into a BW InfoProvider as soon as possible. However, many people don’t know the reason for compression or what the compression job actually does. I will explain compression in detail and show why it is important to the overall health of your BW system.
Compression is an option under the Manage tab of every basic BW InfoCube. To see this option, right-click on an InfoProvider and choose Manage. The compression selections are found in the Collapse tab (Figure 1).
The Collapse tab shows the compression options for an InfoCube
What Does Compression Do?
The goal of compression is to reduce the amount of data that is stored in an InfoProvider. To understand how compression does this, I must first discuss the tables associated with the InfoProvider. Every InfoProvider has two fact tables that store data: the F fact table and the E fact table. To see the tables associated with an InfoProvider, use transaction RSRV (Figure 2). The F fact tables are automatically created and named /BI0/Exxxx and /BI0/Fxxxx, where xxxx corresponds to the InfoCube name.
Would you like to see this full item?