1003-1005

Previous Table of Contents Next

Page 1003

• Discoverer performs SQL operations quickly. Unlike a simple reporting tool, Discoverer will optimize the query path for a user's query and decide whether to use prestored aggregations. When a user asks for aggregate data, Discoverer will check to see if that information exists as presummarized data or not. The user does not need to know how the data cubes are configured or even need to think in terms of data cubes.
• Creating summarized tables with Discoverer is easy. Instead of writing complex GROUP BY statements and creating aggregate tables manually, Discoverer will create aggregate storage structures on the command of a few simple GUI instructions. These new structures use more advanced indexing and compression algorithms and are more efficient than home-spun summary tables.
• Discoverer has a client-side data cube cache. This means that once data is retrieved from the server, a client running Discoverer User Addition will have a ResultsBase cache storing the data on his machine. Complex drilldowns and rotations of the data cube can be executed very quickly because the data is already stored on your computer. The server does not have to go through the sometimes complex task of recalculating aggregates from a large table as the user drills down levels.
• Discoverer offers a query prediction tool called the Predictive Query Governer that will provide the user with the estimated query time before the query is started. Many times users ask for information, having no idea how long it will take the database to return the result set. Discoverer also offers a non-blocking feature, which will run a longer query in the background so the user can perform other tasks .
• Along with Discoverer's Predictive Query Governor, which simply predicts the time it will take a query to execute, Discoverer has a Reactive Query Governor that will allow the administrator to set an upper threshold on query execution time. For example, you could set certain users a 10-minute ceiling on their queries, and using this limit along with the Predictive Query Governor, could save a great deal of time by eliminating long-running queries that tie up system resources.
• Discoverer has very powerful reporting and analysis tools. Users can easily build complex filters with a simple drag-and-drop capability. Discoverer also offers many different kinds of printed reports including tabular, break tabular, matrix, and master-detail, with font and color control. Discoverer also integrates with Microsoft's OLE2 standard to allow this data to populate more common PC software.
• Discoverer implements Hyper-Drill, which enables users to drill down from summary data to the detail rows quickly using a graphical interface. Because Discoverer is manipulating a data cube, the lengthy recalculations, as we move through different summary levels, do not need to be performed.

Page 1004

• Discoverer hides the complexity of your relational database and simplifies complex column relationships into a business area. A business area is composed of folders, which in turn point to tables or views in your relational database. (See Figure 43.6.)
• For instance, you might want to represent a billing address as a simple element to your analysts. With Discoverer you can define the address as being composed of a number of actual data columns :

   billing_address = street_addr1  street_addr2  street_addr3  City ','Statezip_code; 

  With this power, business analysts don't need to be bothered with a specific database design or a number of different designs when they are working with data. They also don't need to know detailed implementations of SQL such as the above concatenation operator.

Figure 43.6.
Discoverer maps your relational database into simple business folders.

• If you want an interface to Web publishing, Discoverer will output any reports or graphical printouts as HTML (Hypertext Markup Language). This will allow you to put valuable analysis information on a Web-based system for general access, or on a company-wide intranet using the new NCA (Network Computer Architecture) paradigm.
• Discoverer is portable. Like most Oracle products, Discoverer is not as platform-limited as other competing multidimensional databases. When you buy this technology, you are dealing with a company, Oracle, which probably has the most portable software of any vendor in the market.

Page 1005

Summary

The key to success in business today is information. Information has been defined as valuable a commodity as any physical capital that a company owns. By having a tool like Discoverer, you can analyze your complex business data that might not be in an accessible form now. In our sales example in this chapter, the company that used Discoverer would be able to quickly find out about revenue changes in relation to price changes. This company could also analyze prices in different geographical areas and possibly price their shoes differently in different regions of the country.

With economic markets being so volatile as we approach the end of the millennium , the companies that can quickly deduce new patterns in their data will be able to move faster to the changing trends in their industry. With a large investment in information systems, a corporation would be foolish not to invest a fraction of that for a tool that will allow executives and analysts to quickly view that data for trends and overall summaries of how their business is doing.

Aside from the business analysis gains of Discoverer, the tool will help companies save money on labor. I remember once working for an insurance company on a huge aggregated report that the president wanted. I spent two months creating various forms of the same report, which took hours to run even after I had spent a great deal of time optimizing the aggregations. Instead of paying for this unnecessary time, the company could have bought Discoverer and, after a short learning curve, could have had a series of data cubes that the president himself could view or drill down with. Reports on the aggregations would not have tied up resources or taken nearly as long to develop. Furthermore, much of what I did was to convert a series of complex columns into a simple business elements. With Discoverer, this process would have been defined and encapsulated so that another programmer a year later wouldn't have to gather the same knowledge and reinvent the wheel.

Discoverer addresses this tedious aspect of aggregating data and collecting it into meaningful business elements. Consider how much your data-processing department spends to generate these types of reports or how much they spent in building a simple online system to do this. After this, consider the processing time it takes to run these reports either in the day or in batch mode. You will probably come to the conclusion that you could save a great deal of money by purchasing an OLAP tool such as Discoverer.

Previous Table of Contents Next