Data Storage Alternatives

Database or workspace?

The types of analyses performed by your application determine the best choice of a data repository. You must examine the benefits of each storage method in light of your application and decide which one most closely matches your requirements. You can choose to store the data for your business analysis applications from these alternatives:

• Entirely in the relational database. Data is stored entirely in a data warehouse and made available to Java applications by OLAP metadata as described in Chapter 4. During user sessions, data is selected and manipulated in the relational database. This method is typically called Relational OLAP or ROLAP.

• Entirely in the multidimensional analytic workspace. As a routine maintenance task, data is loaded into dimensions and variables in the workspace from one or more sources (including the relational database and flat files) and saved for use by all sessions. During user sessions, data is selected and manipulated in the analytic workspace. This method is typically called Multidimensional OLAP or MOLAP.

• Distributed between the relational database and the analytic workspace. The implementation of this model can, of course, vary widely since it encompasses any scheme that draws on the other two methods. A distributed solution may be desirable when an application requires the advanced calculation capabilities and speed of a MOLAP solution combined with the efficient storage of a relational database. This method is typically called Hybrid OLAP or HOLAP.

Related information

If you are interested in a MOLAP or HOLAP solution, then refer to Oracle9i OLAP Services Developer's Guide to the OLAP SPL for additional information.

Query and reporting applications

The relational database is the preferred data repository for most query and reporting applications that require read-only access to the data. For these applications, the relational database offers scalability in supporting very large data sets efficiently and manageability with a single set of administrative tools.

Predictive analyses applications

Analytic workspaces should be used as a persistent data store for applications that support predictive analysis functions, such as models and forecasts, and what-if scenarios. Other design choices, such as the types of hierarchies and use of non-additive aggregation methods, may make the analytic workspace the preferred data repository.

Aggregate data

In addition to where you store the data is the decision of what to store. Aggregate data can either be generated and stored as a data maintenance step, or aggregated whenever it is needed in response to a query. If the data is already available, then response time is quicker. However, you will need more storage space to accommodate the additional data.

In deciding what to store, you should consider the merits of several factors: how frequently the aggregate data is needed; how much time is required to generate the data; whether the response time for generating the data falls within acceptable limits; how much disk space is required to store the data; and whether you have the option of acquiring additional storage media.