Real-Time Data Integration for Operational Data Warehousing
By Irem Radzik on May 20, 2011
Gone are the days when data warehouses were just for reporting, strategic analytics, and forecasting. Today, more companies are using their data warehouses for operational decision making – and thus more critical to the business. To be able to influence operational decisions the analytical environment needs to be able to stay current with the business events happening right now. Therefore an important requirement is to enable lowest possible latency in which new data is delivered to the data warehouse, ideally in real time.
For example, when a snowstorm hits a certain area, the operational data warehouse can help monitor snow shovel sales and then provide information to help determine whether other stores in the affected areas should move more merchandise to the shelves and also offer other related products at a discounted price. There is a lot less value in reacting to data for a snowstorm that occurred 24 hours ago – or even 6 hours ago.
There are many data integration technologies that serve the data acquisition needs of a data warehouse, however only a few offer real-time data delivery with no impact on source systems’ performance. The challenge for the IT group is to determine what solution or combination of data integration solutions will meet their data delivery and performance needs to help propel the move to operational data warehousing. Such data integration evaluations are aided with the understanding of:
- Selection Criteria – This should include considerations for acceptable latency, data quantity/volumes, data integrity, transformation requirements, and processing overhead/impact on availability.
- “Right Time” vs. "Real-Time" – When evaluating solutions, the technology should deliver real-time data capabilities and let the user choose the “right time” as a business decision. Right-time should be a component of decision latency – a user preference, not a technical constraint.
- Transformations – As
the data warehouse approaches real time, transformations ideally should take
place within the data warehouse in order to reduce data and analysis latency.
This eliminates the need for additional steps for aggregating changed data on a
middle-tier server until it is batch processed-- not to mention the TCO savings of not acquiring or maintaining a middle-tier transformation server.
Oracle offers a complete and certified data integration solution for implementing operational data warehouse on Oracle Exadata. Oracle GoldenGate provides low-impact, real-time change data capture and delivery, while Oracle Data Integrator EE provides high performance transformations within Oracle Exadata. Oracle Data Integrator also offers integrated solution for data profiling and data quality to enable analysis with trusted data. Here is a great customer example how Oracle’s products enable the move to operational data warehousing. You can read more about Oracle's data integration solution for operational data warehousing in our white paper. If you would like to read more about how to use Oracle Data Integration products for Oracle Exadata please check out our recent data sheet.