Data warehouse architecture is being influenced by business practices and goals that continue to evolve, notes Russom. The reason: a well-aligned data warehouse reflects the business it serves.
Here are some of the main business drivers of today’s evolving data warehouse architectures, according to Russom:
- Reporting is increasingly important to business operations: Reports are the primary way businesses distribute and consume information as well as monitor operations every day. That means businesses must protect their traditional data warehouse architectures, which are mainly aimed at providing data for reports as well as management dashboards, performance management, and online analytic processing.
- Discovery analytics is to new business development what reporting is to established operations: Discovery-oriented analytics help companies discover new facts, trends, patterns, outliers, etc. about their business, customers, partners, and the competitive landscape. That is why businesses are extending their average report-oriented data warehouse environments by adding new standalone data platforms that better help with discovery analytics—such as columnar databases, data appliances, NoSQL databases, and Hadoop.
- Big Data isn’t about the “bigness,” but rather about business analytics: The ideal way to get business value out of Big Data is through analytics. Therefore, satisfying data requirements of business analytics (either with Big Data or traditional enterprise data) is the “leading driver for change in data warehouse architectures today.”
- Because each department has different requirements, they usually build their own “shadow programs” for BI and analytics: To prevent the systems in each department from becoming data silos, data warehouse architectures are becoming more federated. Several databases appear to function as a single entity and all the data from multiple sources is presented as if it were stored in one place. This enables the architectural plan to extend across different systems in different departments.
- Increasingly, businesses need access to real-time data: The leading edge now is event processing. Instead of storing data to find out what happened or what could have been, businesses need to act on events as they occur. Event processing allows businesses to proact instead of react to risk as well as create opportunities, not chase them. Although traditional data warehouse architectures are designed for “data-at-rest,” real-time capabilities for “data-in-motion” can retrofit into the architecture.
