Data Warehousing
In recent
years, there has been a realization that software systems and data storage /
retrieval systems designed to support the basic operational business functions
do not satisfy the needs of more strategic decision-making. To support these
more strategic and decision-making needs, companies have recognized the need for
maintaining a Data Warehouse or a historical data
store.
OLTP Heritage
In the
1970’s and 80’s, Relational Database Management System (RDBMS) technology
evolved to satisfy business operational needs. Those
needs were for data management systems that provided
high performance access, storage and updating of operational data. The focus was upon On-Line Transactional
Processing (OLTP) systems. With the OLTP focus, these systems were optimized for
retrieving and storing many small pieces of information such as checking account balances, deposits
and withdrawal accessed from ATM machines.
Each of the RDBMS vendors (Oracle, Informix, Sybase, etc.) implemented their systems to be accessed
using a common syntax, the Structured Query Language (SQL). The adoption of SQL to manage these
systems helped to increase their use by providing a common body of knowledge for the limited number
of Database Administrators (DBA) and developers.
Strategic Focus
As the operational software systems matured, decision makers, managers and analysts tried to use these
systems to gain a better understanding of their company, industry, markets and customers. Instead of
wanting to find the account balances for a particular customer, they wanted to look at those customers
whose balances were most profitable for the company. They wanted to query the data to find trends
and relationships between decisions made and results obtained. Instead of looking at small quantities
of data very frequently, they wanted to look at large quantities of data relatively infrequently.
This was a completely different type of need than OLTP.
To accomplish these analytical tasks, the data needed to be organized differently. Instead of needing
to know the current balance of an account we want to know the balance for each day over the last year
as well as the pattern of deposits and withdrawals. We wanted a historical record of information.
People began to think of these historical data repositories as warehouses of data. Hence the term
Data Warehouse.
Data Warehouse Architecture
As the need for data warehousing matured, companies began to develop ways to organize data within
the various RDBMS architectures to enhance their ability to access the larger amount of data required
for analytical purposes. The Star Schema became the primary data organization scheme for Data Warehouses.
In addition, new applications designed to access data from Data Warehouses were developed. Companies
like Cubularity’s parent created applications that accessed Star schema-organized Data Warehouses to
provide OLAP (On-Line Analytical Processing) functionality. Some of these companies created a
proprietary ROLAP (Relational OLAP) approach to provide this functionality for Data Warehouses.
These proprietary approaches each had different access languages and APIs
pplication Programming Interface) making integration difficult.
A New Standard
In 1998, Microsoft created its OLAP engine (now called
Analysis Services)
which could take data from a Data Warehouse and convert it into a physical and logical
multi-dimensional database model for more powerful, flexible, and efficient data analysis.
In doing so, Microsoft proposed and created an
industry standard for accessing OLAP data, OLE Database for OLAP (ODBO). This has been the most
significant innovation for data warehousing and OLAP in recent years.
The monetary, time, and knowledge investment companies have spent toward building Data Warehouses
can be leveraged with client software products based on the new OLAP standard. Products such as
Knowledge Platform
provide low-cost, powerful, and efficient business intelligence solutions for
decision makers, managers, and analysts.
More
Knowledge Platform
Cubularity
Microsoft Analysis Services
|
 |
