Intelligent Enterprise

Better Insight for Business Decisions

RSS
Webcasts
Whitepapers
Subscribe
Home

September 17, 2003

Real Time, Realistically

Your definition of real time has deep ramifications for business processes and the technologies that support them

by Philip Russom

Continued from Page 1

Factors Influencing Real Time

A realistic definition of real time will take into account a few influential factors.

Technical limitations. Temper your definition of real time by considering the technical limits of current state-of-the-art integration technologies and best practices. For example, the closer an integration process comes to real time, the smaller the amount of information that can be copied and processed. Conversely, large data sets require a latent integration cycle. Also, moving information from one system to another increases loads on those systems and the network between them. When production applications and databases carry heavy loads, the constant interoperability that real-time integration assumes may not be desirable or feasible.

Human factors. Few businesspeople need nonstop information. For example, many of us sync our personal digital assistants at the beginning and end of the work day, without continuously updating via a wireless connection. There's little point in frequently refreshing the data behind most management dashboards because most managers haven't the time or desire to view it more than once daily. When accelerating a business process requires some form of process reengineering, expect the human factors to be more challenging and risky than the technical ones.

Business requirements. Implementing real-time integration is pointless if the business processes involved don't need to react in real time. Believe it or not, some companies have integration technology that can operate faster than the company can. When business moves at a relatively slow pace — like in strip mining, rail freight, and sewage treatment — turning on a dime has limited appeal. Slow-paced processes deserve just as much monitoring, measurement, and recording as fast ones, but not nearly as much immediate reaction to change, unanticipated events, or substandard performance.

Cost factors. Many organizations reduce their requirements for real-time integration once they realize its cost and complexity. For example, some business scenarios require multiple complementary integration technologies. An emerging practice couples enterprise application integration for real-time messages and transactions with extract, transform, and load processes for latent movement of large data sets. Double capabilities usually mean double the costs for software licenses, software maintenance, training, and server hardware. In fact, costs can triple when integrating the integration technologies is complex. Even so, the cost may be offset by the benefits of quickly executing time-sensitive tasks, such as customer cross-sell, overnight fulfillment, and just-in-time inventory.

Define Real Time Up-Front

When it's clear that a business process would benefit from some form of real-time integration, define real time specifically for that process and the IT project that will support it. Create the definition in the project's early planning stages. To ensure that your definition of real time is realistic — that is, it fits the business process and is technically feasible — adjust it according to influential factors such as technical limitations, human factors, business requirements, and costs.

Rate This Article

Comments:

Optional e-mail address:

When defining real time for a specific project, describe how close to real time an integration technology must come or (inversely) how much latency the business process can tolerate. This description will influence the type of integration technology you select. Be open to the possibility that latent forms of integration (such as batch-oriented data movement) may suffice, as these technologies tend to be less costly and risky. In most cases, real-time integration isn't a replacement. It complements older, more latent forms of integration, which must remain in place. Some applications require both real-time and latent integration technologies, which increases complexity and cost.

Philip Russom [www.philiprussom.com] is a Giga analyst at Forrester Research Inc., where he provides advice to user organizations about business intelligence, data warehousing, and data integration.

RESOURCES

Related Article at IntelligentEnterprise.com:

"Real Time: Get Real, Part II," June 30, 2003

"Real Time: Get Real," June 17, 2003

"Data as It Happens," May 31, 2003

"A Virtual Point of View," April 5, 2003

"When Time Is Money," April 5, 2003