Introduction to SQL Server 2000 Analysis Services: Working with the Cube Editor - Page 7
November 11, 2002
Let's process our new
cube, and then take a look at the results via the Preview pane on the Data
Select Tools from the
main menu. 2.
Click Process Cube, as
Let's process our new cube, and then take a look at the results via the Preview pane on the Data tab.
1. Select Tools from the main menu.
2. Click Process Cube, as shown below.
Illustration 21: Select Tools -> Process Cube
NOTE: We might have alternatively chosen the Process Cube button depicted in Illustration 22. We might also have right-clicked the cube, and selected Process Cube from the context menu.
Illustration 22: Process Cube Button
The Save the Cube dialog appears as shown below.
Illustration 23: The Save the Cube Dialog
The Cube: New Name dialog appears.
3. Name the cube ANSYS05, as shown in Illustration 24.
Illustration 24: The Cube: New Name Dialog
4. Click OK.
A message box (shown below) appears, warning that the cube has no aggregates designed, and asking if we want to design aggregations prior to processing the cube.
Illustration 25: Warning Message Box: No Aggregates Designed
5. Click No.
The Process a Cube dialog appears, as shown in Illustration 26. The Full Process option is selected by default, as this is the first time the cube has been proposed for processing.
Illustration 26: The Process a Cube Dialog
6. Click OK.
Cube processing begins, as evidenced in the status display of the Process dialog shown below:
Illustration 27: The Process Dialog, Displaying Status
The Process dialog displays status events within the cube processing evolution and then displays a "Processing Completed Successfully" statement in green at the bottom of the dialog when processing is finished.
7. Click Close.
8. Click the Data tab.
We see the actual data presented in the Preview pane, as shown below.
Illustration 28: Partial View, Actual Data in the Preview Pane (Compressed)
Note that our dimensions, Calendar.Time, Customer and Product, are represented by numbers (the ID fields in the fact table) within the various views we can access. This is expected, as we have used only the ID keys to create the dimensions so as to limit the initial cube build to a central fact table.
We must process the cube before browsing actual data anytime we build a new cube and design its storage options and aggregations, or anytime we change a cube's structure (measures, dimensions, and so on) and save the changes to the cube. We must also process a cube anytime we change the structure of a shared dimension (using the Dimension Wizard) used in the cube. If data in the data source (i.e. data warehouse) supporting the cube has been added or changed, processing is appropriate to furnish updated, accurate results when browsing the cube.