IBM Cognos Dynamic Cubes

20 IBM Cognos Dynamic Cubes

Figure 2-15 shows assigning a trigger to a cube.

Figure 2-15 Assigning a trigger to a cube

The intent of a trigger is to run one or more reports, analyses, or both, to preload data cache

of a cube, ensuring that the data is available in the cache for subsequent users.

Several operations can be done against a cube after it is running:

򐂰 Stop

򐂰 Restart (stop, start)

򐂰 Refresh member cache

򐂰 Refresh data cache

򐂰 Refresh security

򐂰 Clear workload log

򐂰 View recent messages

Figure 2-16 shows the menu for administering a cube.

Figure 2-16 Menu for administering a cube

2.1.6 Virtual cubes

Cognos Dynamic Cubes supports the definition of virtual cubes, which are defined in Cognos

Cube Designer as the merge of two cubes, either of which may be a dynamic (base) cube or

another virtual cube. In this manner it is possible to model multi-fact cubes that span multiple

fact tables and business areas of interest.

Chapter 2. IBM Cognos Dynamic Cubes architecture 21

Figure 2-17 shows the organization of virtual cubes.

Figure 2-17 Organization of virtual cubes

With virtual cubes, dimensions, hierarchies, levels, members, properties, and measures from

two cubes can be merged into a single cube, either automatically or manually on an

object-by-object basis.

During query execution, queries that are posed to virtual cubes are decomposed into queries

that are posed to each underlying cube and their results are merged by using a specified

merge rule. Queries for data are posed to cubes only as appropriate. For example, if a

measure is present in only a single cube, only that cube is queried for its value.

Sales In v e nt o r y

Sal es

Inventory

Multiple fact tables.

Multiple ‘base’ dynamic

cubes.

‘Base’ cubes merge d into

a single virtual cube.