The relentless pursuit of business agility in the 21^st century is forcing IT to evolve to higher plains of existence. IT evolution is all about disruptive technologies and best-of-breed ideas merging together to lead to even better solutions for the enterprise. Two such 800 pound gorillas that are coming together presently are Real-Time Infrastructure (RTI) and Service-Oriented Architecture (SOA).

Re-vitalizing your Infrastructure with RTI

Gone are the days when IT infrastructure (storage, network, and servers) was a static and lifeless entity. Today’s agile business models mandate an infrastructure that is equally agile; one that can quickly adapt to the dynamic and evolving business needs, scale both horizontally as well as vertically, and most importantly maintain a high benefit/cost ratio. Our first gorilla, RTI, is the industry’s answer to this business need and has been well adopted across the board albeit with varying names such as Utility Computing, Adaptive Enterprise (HP), On Demand (IBM), Scaleable Enterprise (Dell), Organic IT (Forester), Dynamic IT (IDC), and RTI (Gartner and Unisys).

The following figure shows the conceptual view of an RTI scenario.

The key to successful RTI lies in two primary technological concepts: Virtualization and Orchestration. Virtualization is the abstraction of the physical hardware from the underlying software (operating system in a server or database in storage) to allow multiple “virtual” machines to reside on a single physical machine with the three-fold goal of consolidating servers/storage/network resources, eliminating waste or excesses, and optimizing the usage of each physical machine. Orchestration is the capability of intelligently managing the virtualized environment based on monitoring key business events in real-time (think Business Activity Monitoring or BAM). These events are filtered, processed, and pattern matched with pre-defined rules and policies that dictate how the virtualized environment needs to tweaked to better respond to present conditions. These pre-defined rules and policies form the basis for the “orchestration intelligence” and are directly tied to business requirements and key performance indices. Orchestration is a complex process that involves such things as controlling the lifecycle of virtual machines, allocating available resources, and deciding which business applications get priority over others in case of conflicts or high demand periods. Truly effective orchestration, as you might guess, is heavily influenced by the applications in the business application layer above it.  Application silos that are tied to their deployed platforms and lack the flexibility to leverage the full potential of such an environment can severely limit the effectiveness of a consolidated, virtualized RTI environment. Similarly, monolithic or tightly coupled applications can also lead to inefficient usage of virtualized resources. The root cause for these limitations is an “impedance mismatch” between the existing business application layer (non virtualized and static) and the re-vitalized IT infrastructure layer (virtualized and dynamic). To fix this impedance mismatch, we need to move on to our next gorilla…SOA.

Transforming the Business Application Layer with SOA