It is easy to see how valuable the drillthrough capability will be for other general analysis of our business results, forecasts and so forth, for finance and accounting, human resources, inventory, and a host of other perspectives within the enterprise. Enabling drillthrough with Analysis Services gives us the best of both worlds - summary level analysis, with the ad hoc capability to see the details that underlie a value that warrants further scrutiny. A simple double click of a value in the Cube Browser delivers the filtered list of transactions for examination.
To gain a sense of comfort for the accuracy and completeness of our results, we have only to examine the data within the independent source database. We will do so with the following procedure. Leaving Analysis Manager "parked" where we are:
8. Start the MSSQL Server SQL Query Analyzer.
The Connect to SQL Server dialog appears, as shown in Illustration 44.
Illustration 44: The Connect to SQL Server Dialog
9. Complete the dialog as appropriate to your local environment.
10. Click OK.
A blank, untitled query window appears, pointed to the default database. As most of us are aware, the Query Editor therein provides a convenient point for creating queries (and other SQL scripts), and executing them against SQL Server databases.
11. Using the Database Selector atop the query window, select the WebTrafficAnalysis_DB as shown in Illustration 45.
Illustration 45: Select the WebTrafficAnalysis_DB
12. Type the following simple query into the Editor Pane:
WHERE IPAdd = '18.104.22.168' and Date = '27/Aug/2002'
Here we are asking for the details for the IP address and date upon which we drilled earlier, but this time from the source database, as a means of verifying the results we obtained in the Analysis Manager / drillthrough scenarios above.
13. Select Query -> Execute to run the query.
The results appear (grid tab), as shown in Illustration 46.
Illustration 46: Query Results
We see that, indeed, there were six interactions from the selected IP Address on the date we specified. We can verify virtually any drillthrough details we obtained in this manner to ascertain that we have designed our cube properly from the associated perspectives.
14. Close the Query Analyzer when desired.
15. Close the Drillthrough Data viewer in Analysis Manager.
16. Click Close on the Cube Browser to close it.
17. Exit the Analysis Manager console as desired.
Our lesson concludes with the above actions, leaving us with an idea of an approach for creating a Web Site Traffic Analysis Cube. While a far more elaborate cube can be developed, this will perhaps serve as a general overview of some of the basic considerations that we encounter in this, and similar, real-world scenarios.
In this two-part lesson, Build a Web Site Traffic Analysis Cube, we focused on the design and construction of a basic Web Site Traffic Analysis Cube. In Part I, we briefly discussed potential business reasons for collecting web site traffic data. We then discussed the use of the Server Access Log as a source of data surrounding web site activity, and designed and built an extract package using Data Transformation Services, to illustrate a straightforward approach for extracting and transforming simple data, then loading it into an MSSQL Server data source.
In this part of the lesson, we entered the design and build phases for our Web Site Traffic Analysis Cube. We created an Analysis Services database, and then we connected the database to the relational source table we created in Part I. After we generated the cube, we performed browses of the data it contained, to examine the results of our handiwork. Finally, we verified drillthrough results we obtained through an independent query against the source relational data in MSSQL Server.
» See All Articles by Columnist William E. Pearson, III
Discuss this article in the MSSQL Server 2000 Analysis Services and MDX Topics Forum.