Appistry weaves "fabric" for fault-tolerant apps

Throw a frog into a pot of boiling water, goes the old saw, and it'll jump right out. But drop him in while the water is cold, turn up the heat very slowly, and he'll never know he's being boiled.

Although it turns out not to be true (according to Dr. Victor Hutchison, a professor from the University of Oklahoma), that story is an apt metaphor for the way developers have come to accept a very high level of complexity, says Mills Ripley, marketing manager at Appistry.

"We as an industry have accepted a level of complexity that we shouldn't have, simply because we've come to believe that it's unavoidable," Ripley says. "We should be saying, no, this isn't good enough. We don't want to kill the frog."

Appistry (formerly Tsunami Research) has taken on the complexity challenge at the application development level with its flagship product, Enterprise Application Fabric.

"The industry has traditionally been dominated by an infrastructure-centric approach," says Jason Wyrick, Appistry technical lead, "which relies on layer upon layer of complex hardware to provide a reliable environment for the application. With an application fabric, it's the application, not the infrastructure, that is most important.

Gartner analyst Massimo Pezzini has identified application "fabrics" as a type of grid-inspired enterprise applications server.

"IT organizations are struggling to identify simple, dependable and flexible platforms that support the implementation of business-critical applications based on service-oriented architecture," Pezzini says. "Grid-based enterprise application servers offer a promising approach to this challenge as they provide a high-availability, fault-tolerant software platform atop a fabric of commodity servers."

A "fabric," as Appistry defines it, is an application execution environment designed to free applications--and in turn architects, developers and administrators--from the constraints of traditional infrastructure-centric IT.

The software accomplishes this, Wyrick explains, by managing an underlying pool of servers on behalf of the application. These computers can range from high-end SMP servers to commodity-class machines.

The result is what Appistry calls fault-tolerant applications; the environment in which the applications run are self-healing, designed to survive and recover from infrastructure failure automatically.

Appistry is promoting EAF as an ideal environment for SOAs, particularly organizations that develop, deploy and maintain service-oriented and/or transactional applications, such as those found in the telecommunications, financial services, intelligence and retail industries.

The latest version of the product, EAF 2.3, is now available. For more information, go to:

About the Author

John K. Waters is a freelance writer based in Silicon Valley. He can be reached at [email protected].