Point-to-point integration can be a quick fix for getting information from system to system. All it takes is a bit of code and voilà – connection complete!
But what happens when the programmer who wrote the script falls off the face of the planet? Inevitably, a change will be submitted by the business. Just how long will it take to find the keys that unlock the black box?
A single instance might be manageable but if you’re like most organizations, dozens of hard-coded, point-to-point interfaces run beneath the surface of application development projects.
And chances are, developers used different technologies over the years, making it difficult to know where issues lie, let alone what could be causing a problem.
It could take an army to overcome a situation like this on a larger scale, not to mention escalated cost. So the question becomes: how much of your business’ agility and resources are you willing to put at risk?
- Remain Open: Standard interfaces (which remove logic from the application layer) will enable you to expose the business functionality of applications much more easily without having to understand the complexity of integration logic.
- Keep it Simple: Model-driven interfaces (EDI) providing visibility into application logic will give you a clear understanding of connectivity between your systems and applications. Model-driven interfaces also make it easier to incorporate or remove components, increasing flexibility of the architecture overall.
- Get on the Bus (not under it): Leveraging a bus-based architecture, you can manage connections with applications through one interface. And when the time comes for inevitable system upgrades, you have only have one interface to maintain…one that scales more easily and delivers higher availability.