Data Aggregation and Redundancy Oracle Tips by Burleson Consulting

The Data Warehouse Development Life Cycle

Aggregating Data For The Oracle Warehouse

Data Aggregation And Data Redundancy
As we have emphasized, the pre-aggregation of data to achieve fast response times is a form of redundancy, and we must always be careful to re-compute any summary data when new information is loaded into our Oracle warehouse. To use a very simple example, the computation gross_pay = hrs_worked * payrate computes a gross_pay value that could be stored in an Oracle table. Because gross_pay is computed from the values of other data columns, it is redundant and must be re-computed each time a component values changes.

Pre-calculation of aggregates speeds up the query, and because data is static, there is no problem with ever needing to update the aggregate table. But there can be a problem with increased system overhead when overall averages need to be re-computed, since overall averages will change with each new entry into the detailed fact table. This problem of having to re-computing running averages can be easily overcome by keeping total values and doing the division to get the average at runtime. For example, an average sales amount computation uses two other summary values: number_of_sales and total_sales_amount. Because the calculation of average sales amount involves a quick division of the other factors, it would make more sense to keep the parts of the average and compute the average sales amount at runtime.

It is important to note that defining aggregation of a multidimensional database is no different than defining aggregate tables in a relational database. At load time, the database will still need to compute the aggregate values. MDDBs also employ the concept of sparse data. Because data is aggregated and pre-sliced, some cells on a cube may not contain data. For example, consider a cube that tracks sales of items across a large company. The cells representing sales of thermal underwear would be null for Hawaii, while the sales of surfboards in Wyoming would also be null. Nearly all of the product offerings are able to maintain a mechanism for compressing out these types of null values.

---

This is an excerpt from "High Performance Data Warehousing".

If you like Oracle tuning, you may enjoy the book Oracle Tuning: The Definitive Reference , with over 900 pages of BC's favorite tuning tips & scripts.  You can buy it directly from the publisher and save 30%, and get instant access to the code depot of Oracle tuning scripts.