IT groups with legacy investments are constantly being asked how they plan to make them compatible with next-gen environments. After all, employees everywhere are clamoring for anytime/anywhere mobile access to key corporate information and services. Companies without an effective web strategy are missing an opportunity to engage with customers and create new revenue streams. And, the cloud offers a chance to streamline IT models and lower infrastructure costs.
At some point, every IT shop has to decide – buy something new, recreate, migrate or integrate?
The proponents of integration value the high reliability of mainframe technology above all else. This group can overlook the difficulties associated with trying to add new features and/or integrate with next-gen mobile, virtual and cloud environments. On the other hand, proponents of migration tend to prioritize the fast path to new features and new accessibility options over uptime. The time to market difficulties of re-writing or buying a new application make them very difficult to rationalize.
In a purely theoretical world, supporting technology shouldn’t actually make a difference. There are open, standards-based programming techniques that allow developers to deploy fully integrated mobile apps in just weeks. And, companies can always boost infrastructure resources to a point where next-gen system can deliver the reliability levels offered by mainframes.
However, no one works in a theoretical IT department – in the real world administrators must also evaluate risk factors. Clearly, every organization must evaluate risk on a case by case basis, however, there are a few general factors to consider. Fully scale infrastructure refreshment can be an extremely difficult, all or nothing project. Because the old and new cannot work together, the project must be 100% successful to be worthwhile. There is no phased approach to this model. While a POC can be used to test the viability of an idea, the environment must be built to scale before IT can be sure that it will adequately support the new functionality.
Although there are risks in any integration project, the effect of failure can be mitigated by working in phases. Companies can start testing concepts with a relatively simple application (say rewriting an existing functionality for mobile access). In this way, the potential cost of a failure and the time invested will be a whole lot lower. Once validated, that “POC” can be implemented straightaway.
Each organization will have to answer the integrate vs. recreate question for themselves. Only they know how well their legacy systems are functioning and how vital they are to the organization. With enough time and money, companies can usually make any technology work in the long run. The real question remains – how much are you willing to risk to get there?