The event took place on thursday the 28th of october. During four presentations Oracle presented Oracle OLAP in the data warehouse, the OLAP architecture, Oracle products (like OWB and discoverer for OLAP), and more recent products/add-ins like Oracle Portal, BI Beans and the Excel add-in.
The event was well attended and took off with a presentation about Oracle OLAP in the data warehouse. The presentation showed on a high level the possibilities with Oracle OLAP in the data warehouse. In particular the presentations about integrating the data warehouse in Oracle Portal (very briefly described though), the BI Beans add-in for JDeveloper and the Excel add-in for, yes, Microsoft Excel was quite interesting!
From the ‘hackers’ event I choose to have a better look at the BI Beans add-in for JDeveloper and the Excel add-in. The BI Beans add-in for JDeveloper (10g) can be downloaded from the Oracle web-site (BI Beans download page). The installation manual can be downloaded from the Oracle web site (installation manual) also. Unfortunately, the Excel add-in comes not with the Oracle 10g enterprise edition but with Oracle OLAP, which needs to be purchased separately.
I will describe the BI Beans functionality as well as the excel add-in shortly. Comments of other colleagues on the event are written down below and perhaps a couple of background documents will become available later.
BI Beans
With BI Beans, Oracle provides functionality within JDeveloper to present data from the data warehouse via a web browser. After creating pages with the Business Intelligence Beans, the project can be deployed. This means that data from the data warehouse can be retrieved via a web browser everywhere around the globe. But of course this type of (delicate) data is more suitable for an intranet application.
With BI Beans you can create tables in where the user is able to down and up-drill or changes the line of approach, although you have web page limitations.
For instance, a user is looking at data in where the total of sales per country is shown. The user then can up-drill by choosing a country and see the total sales per product. By changing the line of approach, the user can choose a product and see the total sales of that product per country, etc.
You can create graphs in where the user is able to down and up-drill by easily click the graph! Changing the line of approach is possible here also, but the original layout of the graph remains the same.
Excel add-in
The excel add-in gives a new menu option in Microsoft Excel with the name OracleOLAP. To make use of the add-in, you have to create a connection with the database. To create a connection with the database you make use of the same wizard as the wizard used to create a connection with the database in JDeveloper. Then, a query wizard, which again is the same as the one used in JDeveloper BI Beans, guides you through the necessary steps. Data is retrieved from the database and presented in Excel (without formulas). You are able to up and down-drill within Excel. By up and or down-drilling, a connection with the database is required because not all the data is present whenever the query is executed the first time. The functionality of Excel can be used, which is an advantage but at the same time a disadvantage. For instance, you can summarize a column in excel which gives a certain value. When you up and or down-drill, the value can change…
Business Inteligence OLAP event, continued…
I was there too. It has been a week ago. What do I remember? That “OLAP is fully integrated in the oracle database” and that … “OLAP is fully integrated in the oracle database”.
Yes, Oracle did a good job. Their main sales pitch is stuck in my head. Of course they also explained that OLAP itself is very important. End Users are more interested in (on line) KPI’s and not so much in raw data. The multi dimensional end user model is easy to understand, etc. But we know that already.
The main message is; you don’t have to store you multi dimensional data outside your main database. You can use Oracle Warehousebuilder to create a multidimensional model and store its physical components in the same database that stores your star model. Oracle Warehouse builder makes it even possible to map your tables on dimensions, measures and cubes.
I had the opportunity to ‘hack’ an OLAP environment, including a mapping between my facts and dimension tables and my multidimensional objects. The beta version of Oracle Warehouse builder was a bit buggy, but it works.
Business Intelligence OLAP event, continued…
Warehouse Builder:
Oracle Warehouse Builder provides a complete environment for the design, population and management of data warehouses. One of the crucial improvements in the coming release is the seamless deployment, execution and data viewing of OLAP objects. Other major benefits of the coming release of Warehouse Builder are:
• All functionality embedded in the common editors you are used to for relational modeling.
• Direct deployment of any object
• Integration within Process Flows and dependency management.
During a short period of time we got an overview of the new functionality. The coming release seems easier to use than the current Warehouse Builder.
Discoverer:
Discoverer is a tool for ad-hoc query, reporting, analysis, that gives users at all levels of the organization access to information from data marts, data warehouses and online transaction processing systems.
The new release contains a functionality to give access to OLAP-cubes. The access is separate from the relational discoverer. Both Discoverers have the same look and feel. The desktop-version of the Discoverer Plus disappeared; it is only available via the web.
We also discovered some advantages for the relational part, which is not yet available in the Discoverer 4-release.
• The use of footnotes and being able to alter the title of workbooks.
• Drilling on cell-level to other workbooks and even to an URL-link.
• The printing-options have improved a lot. For instance fit-to-page options.
• Administrative options like sharing workbooks are more clear and easier to use.
• Nested parameters.
• List of Values can display descriptions and use IDs in background. This will improve the performance of the workbooks.