Intelligent Enterprise

Better Insight for Business Decisions

RSS
Webcasts
Whitepapers
Subscribe
Home

May 9, 2002

An Embarrassment Of Riches

What's next for database technology in the '00s?

by Justin Kestelyn

There's a curious thing about data: You can never get enough of it, yet most businesses have more than they can handle.

This "embarrassment of riches" is one of the several topics addressed in this special data management issue of Intelligent Enterprise. Most notably, we have convened — as far as I know, for the first time in a purely editorial setting — a "round table" comprising leading database performance experts from the major commercial software vendors in the DBMS marketplace today.

As you'll discover in "What's Next for the Database?", these experts — Stephen Brobst of NCR Teradata (a former contributing editor to Intelligent Enterprise, who heroically participated via telephone from India), Don Haderle of IBM, Ken Jacobs of Oracle, and Prof. Jeffrey Ullman of Stanford University — spotlight some of the most important technical challenges facing database technology strategists as they redefine the role of the DBMS in emerging strategic business applications. (I should note that Jim Gray, a Microsoft Research distinguished engineer and one of the inventors of classical database transaction management, was invited to participate but other commitments precluded his appearance. We managed to keep Jim involved, however; read David Stodder's interview in Strategic Knowledge.)

The panelists agree that the traditional technical quandary of the last quarter century — the availability and scalability of batch online transaction processing applications — has mostly been solved. In its place, however, have arisen a slew of new problems that are pushing current architectures to their limits. Here are some highlights from the discussion:

Rapid information integration. The panelists agree: Emerging Web services protocols offer an intriguing potential solution to the challenge of rapidly discovering structured and unstructured data and then integrating it with business applications on a semantic level.
Automated performance management. Although "traditional" availability challenges have been solved, the panelists raise the fact that current approaches (as well as the humans who implement and manage them) are straining under the weight of escalating database size and complexity. New, automated optimization techniques are in order.
Revitalizing SQL. Business needs such as realtime data mining and transparent access to nonrelational data will drive the development of analytic and object/relational extensions to SQL itself. In the panelists' opinion, these trends will keep SQL "vibrant" for years to come.
Database ascendant. The DBMS will continue to expand its functional footprint, ingesting increasingly powerful analytic and data cleansing/data transformation engines. Furthermore, the panelists predict that mixed workloads will eventually become the norm.

It Gets Better

The story doesn't end there. In this issue's other feature stories, two experts in their respective fields, Richard Winter and David Loshin, offer important advice about two pressing issues in intelligent organizations today: scalable Web analytics and enterprise data quality, respectively.

As Winter discusses in "Over the Wall," many organizations that have recognized the business value of e-business analytics are becoming victims of their own success: the more customer behavior information they acquire, the less useful it becomes as performance and scalability limitations take their toll. In this article, Winter offers a methodology for crafting requirements that help you avoid that "brick wall" before it materializes.

Rate This Article

Comments:

Optional e-mail address:

In "The Golden Rules," David Loshin discusses another crucial data management issue: what it means to have a true enterprise data quality strategy. As Loshin points out, the terms "data cleansing" and "data quality" are commonly confused; the former implies a tactical, one-off strategy, while the latter implies an enterprisewide strategy for data integrity. Here the author describes a conceptual enterprisewide approach based on business rules.

Coming Attractions

That's just a taste of the compelling menu we've prepared for you in this issue. In the future, we plan to convene more luminary round tables to address other topics of interest to you. As always, we'll keep you posted.