Smart Enterprise Magazine

Volume 7, Number 1, 2013

Issue link:

Contents of this Issue


Page 5 of 43

Leadership Perspective Virtual simulation and modeling techniques speed development, improve quality and deliver ROI. | By John Michelsen C onstraints are the enemy of speed. Infrastructure, data and systems availability can delay the completion of tasks in the software-development lifecycle. They���re also the main reason why software projects are delivered late, over budget or with poor quality. Fortunately, a solution is at hand. Leading-edge IT departments at several companies ��� including FedEx, Sprint, DirecTV and First Data ��� are overcoming these software-development constraints with modeling and simulation techniques. Both modeling and simulation are tried-and-true in certain industries. Airplane designers, for example, prove and validate every component of a plane with computer modeling and simulation before subjecting the aircraft to the real world. First, the safety and effectiveness of each component ��� the wings, landing gear, tail ��� are proven separately. Then the components are assembled into a safe, working plane. Today���s composite business applications, like the components of an aircraft, are typically interrelated and interdependent in a complex environment. Yet modeling and simulation are rare with business software. While aviation engineers can test an airplane���s wing in a wind tunnel long before it hits the runway, software engineers tend to test applications at integration time, in turn plaguing the business with high levels of unpredictability. The problem? Constraints. The more complex and distributed our business software becomes, the more we suffer the constraints of systems and services that are simply unavailable when we need them. One propertyinsurance CIO put this problem this way: ���I can never do anything until I have everything, and I never have everything!��� Get ���In-Scope��� Here is where the advanced simulation capabilities of service virtualization (SV) software can prove beneficial. Most conventional forms of hardware virtualization focus exclusively on 6 copying the ���in-scope��� components and data, which are the subject of development or test activities. But what about everything that is either unavailable or not yet ready for work? SV lets teams capture and simulate these other dependencies in a virtual environment, eliminating the need to wait. To fully appreciate the benefits of SV, consider a critical new mobile-banking app that needs to be revised. In the past, each team of developers and testers would need to request access to servers, transaction gateways, mainframes and other related distributed systems, all of them lined up with the appropriate customer data. But too often, weeks would pass before the developers finally got shared access. As a result, the project would start late, and the developers would encounter a wide range of constraints, including limited capacity, conflicting delivery schedules, data-management and volatility problems, and third-party costs and control issues. By contrast, SV lets those application developers capture and simulate the behavior, data and performance characteristics of mainframes and distributed systems, as well as those of other dependent elements, including back-office systems and thirdparty services. These simulations then are catalogued as virtual services. So instead of working in a live environment, the developer instead works in a virtual service environment (VSE) that is ready to use on demand, 24x7 and with minimal incremental costs and setup time. Once the developers have a VSE, they can do development, integration and performance testing by interacting in parallel with these virtual services, instead of with the real systems ��� and their many constraints. When it comes to the supply chain of business software, SV is nothing less than transformational. Our customer case studies show that working with SV speeds overall application development by 25 to 50 percent, reduces defects by 80 percent, and produces a time-to-ROI that���s simply amazing. We have dozens of customers saying, essentially, ���I invested a million, and made back more than that, literally in weeks.��� In fact, our fastest recorded ROI is just nine days, and the second fastest, only 20 days. The era of constraints is indeed ending. n John Michelsen is the CTO of CA Technologies and co-author of Service Virtualization: Reality Is Overrated (Apress, 2012). photograph: CA Technologies Simulate Your Way to Faster Time to Market

Articles in this issue

Links on this page

Archives of this issue

view archives of Smart Enterprise Magazine - Volume 7, Number 1, 2013