subject: IT Lifecycle Management – The Semi Circle of Life [print this page] Gaps remain in the IT Lifecycle Management circle, but what are the sector's vendors doing to plug them? What should an ideal-world integrated, automated, end-to-end ITLM solution really look like?
Ah, the three letter acronym so familiar and staple a part of the technology lexicon that we gave it its own er, three letter acronym. The T.L.A. Honestly. Only in IT. It seems only fair and equitable then, that someone should finally stick up for a less popular member of the abbreviation clan the FLA'. For while we seem to have developed a certain cheeky fondness for acronyms of three letters, their four-letter cousins remain for the most part lonely outsiders. Persona non grata whose presence is sniffed at suspiciously by most and discouraged by almost all.
Take ITLM. As technologies go, IT Lifecycle Management makes a good deal more sense than most. It's just that fourth letter isn't it? It's so off-putting. Get past the irritating extra consonant though, and most people are sold. Dubbed IT lifecycle management (ITLM) by IDC, what we're talking about here in a nutshell is the much-vaunted virtuous IT circle a recognisable technology lifecycle wherein each stage of a system's life is clear and transparent and moves fluidly into the next: starting with its planning, through its development, into everyday operation, and thence to its end of life the end game slash holy grail being a fully joined up, automated, end-to-end solution.
In truth however, many gaps still have to be closed before a true circle, much less a virtuous one, can become a reality. Accordingly, many major IT vendors have begun trying to fill in the holes; integrating application lifecycle management and systems management tools, and in turn integrating these utilities with portfolio management tools, web technologies and other platforms.
As Chris Limberger, product manager for Application Availability & Performance at Macro 4 suggests, the need is essentially to tie development and operations (and ultimately the needs of business users) more closely together. Why? Quite simply to maintain control and ensure quality the very real need for which has been neatly illustrated by high profile problems such as those experienced with Heathrow's T5 baggage handling system recently. "The suspicion with outages on new systems like this is that they may not have been managed rigorously enough through development and then into production", he says. Such management involves elements like adequate stress testing, simulating business processes in real world' conditions, performance monitoring and pre-emptive actions.
According to Limberger, one of the biggest gaps in the ITLM loop right now is that whilst many systems exist to monitor performance, they tend to focus on specific bits of the infrastructure, or on particular applications such as the database or the network. The trouble with this, he says, is that problems often occur in the margins between applications, databases, and operating systems.
In other words, enterprises need to take a more holistic view of applications and the entire environment in which they operate. Businesses must, quite literally, learn to read between the lines. "Similarly, development teams and those managing and monitoring systems once they are in operation should be sharing the same tools when identifying and analysing problems", he adds. "Often they rely on different systems, making communication and collaboration on problem resolution difficult."
Another gap exists in the handover process between Business and IT, suggests Oliver Bendig, Senior Product Manager with FrontRange Solutions. He believes it is imperative to ensure that changes to IT are controlled and transparent to the business from application procurement to application packaging, testing, release, execution, and feedback.
Bendig also feels that the high complexity of different systems creates further key gaps, and that businesses therefore need solutions agile enough to support existing processes. "The problem with all embracing, fully integrated Application Lifecycle Management (ALM) solutions is that organisations already have a lot of tooling in place, and rip-and-replace is costly and disruptive," agrees Michael Azoff, senior research analyst with the Butler Group.
Conversely though, Azoff believes that plugging existing gaps with high-end ALM solutions often means companies don't get the full benefit of integration tracking work items across the lifecycle and a whole host of project intelligence that can mine integrated data sources. Indeed, leading ALM vendors such as IBM and Borland are tackling this by creating platforms open to plug-ins from third party tools. "So IBM's jazz.net platform has a growing circle of partners and Borland's latest Borland Management Suite rests on their Open ALM approach", adds Azoff. "But it's early days."
Andy Seager, VP of Global Product Marketing at Borland, adds that software delivery is still not a managed, repeatable process, mostly lacking in the discipline and controls that other business processes now take for granted. As such, he explains, different companies are at various levels of maturity across the lifecycle and have no way of accurately measuring or improving activity.
"What is needed is for companies to adopt a maturity model and a system where they can accurately measure progress and improve iteratively", he says. "Right now, most software teams are flying blind and don't really know why they succeed or fail so most of them fail. What's needed is the critical in-flight information that allows teams to make decisions that lead to successful deliveries."
Borland's Open ALM tools claim to allow businesses to automate, track, and measure key processes in software delivery, enabling them to use any combination of lifecycle tools on whichever platform they choose. The ability to measure and change key processes throughout delivery means you can spot issues before they become problems, ensuring that what's delivered suits changing business needs and ultimately improves the overall delivery of the project.
For Clive Longbottom, service director at analysts Quocirca, an integrated, automated, end-to-end IT lifecycle management solution also needs to include some form of granular discovery capability, at both the IT hardware and software asset level. The solution must also understand the dependencies and context between these assets, he says.
"It then needs to know what the overall business processes are that the technology is there to facilitate, how these break down in to business tasks, and how each task uses the technical functions available."
The next step is then properly auditing usage and effectiveness against business needs. The solution has to be able to pick up on implicit and explicit user needs, identify any redundant steps in a process, and raise development tickets around this, all whilst managing any requested changes, says Longbottom. There's then a need to manage all the patches and updates to applications, services and functions, as well as hardware firmware updates. Within SOA (Service Oriented Architecture) style environments it's also necessary to cater for contract negotiations between calling and serving functions and the provisioning and deprovisioning of physical and virtual resources on a dynamic basis.
