Intelligent Enterprise

Better Insight for Business Decisions

RSS
Webcasts
Digital Library
Subscribe
Home

September 17, 2003

Breaking Out of the Warehouse

To many, BI is equivalent to data warehousing. Although data warehouses are important, they're not the total solution. To deliver all the BI capabilities business users want, we need to create a BI grid

by Michael L. Gonzales

Continued from Page 2

Hooking Up to the Grid

The BI fabric must provide a means to embrace all the architectural layers of BI, accessing directly the specific object of interest and providing a means to act upon any particular event or data, or any combination of events and data.

Network management software (NMS) is a grid technology serving as an excellent foundation for BI applications today. The NMS role is traditionally focused on managing hardware resources for complex networks — providing split-second decisions in areas such as load-balancing system resources to network traffic demands. But this technology is also the necessary infrastructure for an expanded role into the BI domain.

NMS technology is specifically designed to provide cohesive resource management, essentially working in highly diverse and disparate environments, monitoring and reacting to events across multiple and diverse computing platforms, operating systems, and network protocols. Therefore, opening this technology to provide other monitoring and action-based functions is simple. For example, if your NMS already has agents on systems throughout most of your enterprise, then that software will serve as the conduit by which you can add functions for monitoring, collecting, analyzing, and reacting to the data itself and other events that go beyond, for instance, CPU resource availability. Platform Computing Inc. is one such vendor exploiting its existing NMS technology to facilitate a broader role in BI applications.

Of course other companies that don't have an NMS genealogy are also attempting to supply the BI grid. For example, SeeRun Corp. has a real-time event monitoring and response system based on process models. This software allows companies to model specific processes — for example, order fulfillment — and monitor the necessary events of that process to ensure efficiency and to head off potential bottlenecks or other problems.

Stop the Gap

BI is never manufactured from a single piece of data or a particular event. Instead, BI comprises several disparate data pieces and events that, when taken as a whole, create intelligence. Think of an image displayed on a computer screen. The image manifests only by the synchronized effort of hundreds of pixels, each representing a single piece of data. Although the pixel is critical to create the overall image, it's the image itself that affords value.

Rate This Article

Comments:

Optional e-mail address:

The most difficult challenge in delivering BI is to create, reliably and on demand, this amalgamation and correlation of data and events in order to produce a BI image that fulfills a specific business need. This difficulty comes mainly from the fact that traditional technology and techniques used in BI efforts are focused on a particular task, such as extract, transform, load or OLAP. These warehouse-centric technologies simply aren't designed to reach across the enterprise and — in real time — monitor, collect, and analyze an array of data and events.

However, just because warehouse-centric technology can't fill the BI gap doesn't mean that no technology exists to do so. Mature applications born in the IT management arena have a proven record of monitoring and collecting highly disparate data and events from complex networks, systems, and applications. Moreover, the best of these IT management applications can create, on demand and in real time, composite and correlated views that facilitate better decision-making.

Michael L. Gonzales [mlg@starfocus.com] is the president of The Focus Group Ltd., a consulting firm specializing in data warehousing. He has written several books, including IBM Data Warehousing (Wiley, 2003), and speaks frequently at industry user conferences, and conducts data warehouse courses internationally.