With so many major IT outages occurring this year, organisations are realising the importance of tracking the software quality and risk metrics of their core business applications to keep them robust and reliable.
It’s proven difficult for most practitioners to collect relevant risk metrics, and even more difficult to put them in context. That changes today with the introduction of a new capability for CIOs to benchmark the quality of their applications against competitors in their industry, as well as ease the burden of analyzing and measuring the software structure of enterprise-sized applications.
Lev Lesokhin, EVP, strategy and market development at CAST told CBR: "Measuring software is difficult to do. It’s like nailing jelly to the wall. It has not been done in the past quite simply because of how difficult it is to do. Product- gathering source code and reverse engineering needs to be done until you know what it is that needs to be measured. To reconstruct is tough."
Lesokhin spoke of the five elements of software building that need to be embraced: robustness (i.e. stability), performance (i.e. efficiency), security, transferability and changeability (i.e. ability to update and move with the times). "You have to be careful to understand each business requirement and comport changes for functionality as the changes tend to propagate," he said.
This is what the new Application Intelligence Platform: AIP 7.1 from CAST does. It examines these key elements of software building and the reliability of software systems across their industry.
And with the addition of a new Delivery Management Tool, customers will be able to upload their source directly within the CAST AIP portal with the press of a button.
AIP 7.1 comes to market with several important enhancements that ease enterprise-level software analysis, including enhanced configuration management, a new custom reporting engine, an extended API, and advanced source code delivery automation. These new capabilities turn AIP into the closest the market has seen to a fully automated, plug-and-play solution for the enterprise. The updated dashboard — currently in beta — is also available on tablet and mobile devices for executives who require up-to-the-minute application analytics on the go.
"I’m sure the CEOs of NASDAQ, Knight Capital, American Airlines and Sabre wished they had CAST AIP 7.1 pointed at their critical IT systems before they crashed and made headlines, tarnished their image, and sent their stocks plummeting," said Dr. Bill Curtis, chief scientist at CAST commenting on hoe Knight lost around $170m in just 45 minutes because of an IT outage which eventually forced them to recapitalise.
"Every one of these companies could have benefited from having the latest AIP dashboard on the CEO’s desk. I’m confident none of these failures would have happened if they had been able to see this train coming before it went off the rails."
Lesokhin added: "Goldman Sacs also had an issue pricing their options too low, due to a problem with their algorithms. They lost an estimated $100. These companies need efficiency and robustness to avoid these problems. If they do encounter a problem, they need to act fast to rectify the situation."
Lesokhin also touched on the issue of responsibility of re-engineering: "It is becoming increasingly difficult for IT to manage these problems, people are referring to the chief intelligence officer as ‘C. I. No’ because they are being asked to do things that are not possible.
"Nobody is responsible. Y2K gave a chance to reprieve and do massive re-engineering, but now it is 13 years later and there is nobody responsible for the management of these data processes."
The new reporting capability of AIP uses the extended, REST-ful web service API, linking preconfigured templates in MS Word to the AIP knowledgebase. To customise additional reports, the customer simply has to configure a new MS Word template, connect the fields to those available via the API, and hit go.
Lesokhin describes SEM as basically a fancy filing system: "You have to have secure management and need to reverse engineer it to find out where the information is going.
"It is important when trying to measure legacies, such as a big java development from 2005. In this sense a legacy is anything that doesn’t have automated test scripts," he said.
A new level of source code delivery automation is introduced in the Delivery Management Tool (DMT). The DMT plugs into all forms of source code management and build management tooling. Integrations to major tools are available on the CAST Customer Forum, exotic or custom tools are integrated by scripting.
The DMT then orchestrates collection of all source code elements needed to fully represent a snapshot of an application. DMT automatically checks the completeness of the application, ensuring apples-to-apples analyses between snapshots – an issue that’s critical when implementing enterprise-grade software measurement.
In its latest incarnation, AIP provides advanced application analytics for decision support, software risk prevention, and application visibility unavailable to this point at the enterprise level.
