Ready for This Enterprise Strategy?

According to the Gartner Group, the concept they dub "Zero Latency Enterprise Strategy" is one of the most important computing trends of the next five years. Drawing a parallel to the impact of OLTP back in the 1970s, the IT research firm says that a zero-latency enterprise strategy has the potential to achieve similar breakthroughs by applying the notion of immediate communication among multiple, independent distributed systems.

As distributed systems and client/server topologies have slowly assumed more prominence over the last decade, the pace of business activity, or velocity of the business functions, has picked up. This need to act quickly has put strain on the business infrastructure, increased IT workload, and magnified the value of timely information distribution.

Gartner's suggestion that businesses with dispersed operations make zero-latency a strategic principle formalizes the current struggle to ensure that everyone in an enterprise is taking actions based on the same data. The timeliness with which data replication is needed across the enterprise varies with the use to which the data is put. Currency of data should be sufficient to support timely decision making, and if zero-latency strategies become more commonplace as we enter the next century, it may well yield the same kinds of competitive advantages OLTP created two decades ago.

This is not total "pie-in-the-sky" thinking. There are limited real-world applications today by such leading-edge users including 3Com, Bechtel, Daiwa Securities and Delta Air Lines. These companies have re-engineered some of their business processes, tailored some application systems to real-time operation, and implemented a broad messaging infrastructure that attempts to bring more wide-spread data currency where it can yield significant benefit.

While the concept implies no specific technical architecture, it relies on implementing middleware that exploits the immediate exchange of information across technical and organizational boundaries to achieve business benefit, as Gartner defines it. In addition to handling immediate data distribution across different computer platforms and database management systems, a zero-latency environment must also interface with current application systems. Furthermore, since absolute "zero" is not really practical, any implementation that reduces latency, or data "float," to the point at which further reduction lacks real payoff, is acceptable.

At present there are several middleware providers whose software is being used by the early adopters of this strategy. Because the concept involves changes in the way a business reacts to events within it, and since such reactions generally involve decisions and actions by people as well as application systems, implementing this concept is at best "non-trivial." But then it has often been so with potential big-payoff strategies in IT.

While most of the middleware providers are relatively young players, one company has roots that go back 15 years in solving problems of data replication and transfer. ITI Inc. is currently marketing a product named Shadowbase that evolved from meeting the challenge of maintaining real-time database duplicates in OLTP environments running on Tandem systems. From that beginning it has broadened its scope to include data replication between a variety of source and target platforms. The method used by this software is "non-invasive." It does not add read/write cycles to the application processing on the source database. Instead, it watches the "journal file" created by the application and replicates data across the enterprise based on the activity reported in the journal file.

Gartner suggests that the notion of zero latency can be best-understood in the context of other time-related management strategies and can complement some of them. For example, just-in-time (JIT) manufacturing or distribution, which drives down, inventories costs. Or straight-through processing (STP) sometimes used in financial institutions where one transaction entry automatically flows through its "life cycle" to update information in different application systems in various locations.

An ITI Inc. data replication white paper by H. William Knapp, claims that the reduction in costs of computing and data storage coupled with the speed increases in communication make a strategy of data replication important for the business benefits which can result. He predicts that when data replication is employed effectively, companies can reap great gains in the benefits from distributed processing as a result of more effective end-user processing.

As the column title suggests, this strategy may not be for everyone. On the other hand, examining your strategic goals in the light of the concept of zero-latency can either reassure or unnerve you about your future database strategy. Having been raised as a potential "issue" and armed with the knowledge that some are developing a zero-latency strategy, ignoring the idea altogether seems unwise.

After 18 years in marketing and sales at IBM, Bob Diefenbacher founded Denbrook Systems Associates, an IT consulting firm based in Malvern, Pa. [email protected].

Must Read Articles