High resolution satellites, multi-modal sensors, and other input sources are driving an explosion in data available to the Intelligence community. At the same time, today’s global security environment is driving the need to more rapidly turn data into insight and effective action. Processing more data, faster, presents a considerable challenge.
Intelligence agencies cannot afford delays in uncovering actionable insight through data- and compute-intensive applications such as signal intelligence, geospatial intelligence and cryptanalysis. At the same time, agencies and their commercial suppliers must work within operationally constrained environments. Unfortunately, traditional IT approaches simply fall short when it comes to providing a solution that can reliably process massive amounts of data in a timely fashion without overburdening agency budgets and labor resources.
Appistry Enterprise Application Fabric (Appistry EAF) provides a “scale without fail” application environment that enables intelligence agencies to run large-scale data-processing applications more quickly, easily and cost-effectively than previously possible. By building scalability, dependability and manageability into the application layer itself, Appistry EAF insulates mission-critical applications from underlying infrastructure failures, while empowering agencies with new levels of scale, agility and cost-effectiveness to meet the intelligence demands of today and tomorrow.
The Challenge: Failure Is Not An Option
Intelligence agencies face enormous pressure to meet ever-increasing mission demands. With national security at stake, failure is simply not an option. To meet their objectives, agencies rely on technology to help them quickly process massive amounts of data into insights that can be acted upon before a threat becomes a crisis.
However, with threats intensifying at the same time that data sources are expanding, traditional approaches struggle to keep up. If data-to-insight latencies continue to increase, analysis innovations that lead to intelligence breakthroughs may be stifled. The end result will be agencies that are “treading water,” rather than moving forward in new directions that may ultimately make the difference between mission success and failure.
Solution: Appistry Enterprise Application Fabric
Appistry Enterprise Application Fabric is a software-based environ-ment for running time-critical, CPU-and/or data-intensive applications (e.g. geospatial and signal intelligence) on a self-healing network of commodity-grade computers. Appistry EAF creates a next-generation “scale without fail” grid environment by building scalability, dependability and manageability into the application layer. This approach insulates fabric-based applications from the underlying physical infrastructure and its frailties, so intelligence applications can survive inevitable infrastructure failure without loss of data or interruption of service. Overall, unlike traditional grid technology, Appistry EAF enables cost-effective scaling without sacrificing application dependability or manageability.
Appistry EAF: Next-Generation Platform for Intelligence Applications
Appistry EAF software makes it easy to achieve mission objectives by predictably and reliably executing large-scale intelligence applications. Appistry EAF automatically organizes machines to maximize scalability, balance load across the computing infrastructure, and compensate for hardware and software failures. Because Appistry EAF does all of this dynamically, in a fully distributed manner, no single point of failure exists in the application fabric and all work is completed successfully—on time, every time. Furthermore, new computers can come online and existing computers can fail, all without interruption to running applications and without manual operator intervention.
Limitations of Traditional Approaches
Some agencies have looked to traditional technology approaches, including “big iron” and traditional (first-generation) grid computing, to meet their needs for additional data-processing capacity. Unfortunately, these approaches demand serious trade-offs among fault-tolerance, manageability and cost-effectiveness that severely limit agency effectiveness.
“Big iron” vendors have had great success in the past serving the Intelligence community; however, the challenge today is to deliver an application environment that is simultaneously scalable, dependable, manageable, and affordable in order to maximize agencies’ technology and business agility. Meeting that challenge requires a next-generation approach.
Challenges Posed by First-Generation Grid Computing
Manageability Traditional grid computing produces high overhead costs through poor manageability, requiring that each resource in the grid be managed separately rather than as a single, virtual system.Deployability First-generation grids are often difficult to deploy, requiring significant manual intervention to bring resources on line and unwieldy re-architecture to get existing applications up and running.
Predictability While grid computing has proven tremendously valuable for academic pursuits that have comparatively relaxed require-ments, a more manageable, “real-time” grid is required to meet the demanding needs of intelligence agencies.
