Every business has invested countless sums in technology over the years. Since the Harvard Mk1 in the 1940s, organizations have been using technology to help run their business. However, a lot of investment today is in a digital front end — mobile apps, web portals etc — at the expense of the back end, which makes those apps work.
Without a modern infrastructure, ‘digital transformation' is built on sand, and a big reason why seven in ten digital transformation projects fail (per McKinsey). Mainframes, neglected ERP instances or other such systems are holding back modernization when they're siloed, but if they are truly connected, they can be an asset to transformation projects.
Accenture research tells us that 70% of businesses believe that ‘technical debt' holds them back from modernization, but at the same time 70% want to keep their core systems for as long as possible. What that means is that we need a fast lane to modern infrastructure that brings historical investments along for the ride.
Turning legacy into bedrock
To many non-technical folks — and perhaps even to many technical people too — you mention things like mainframes or legacy ERP and they will groan. The 70% of people who said that these systems stifle modernization efforts aren't wrong. Because ‘old' often means ‘siloed' — and that's the nature of the problem. The tech itself, and certainly the data within it, could actually be a fundamental part of a modern organization. It just needs to be put at the heart of a truly connected enterprise.
In our Adabas & Natural team we talk about "freedom for legacy." We believe that running more of your enterprise applications in a hybrid application landscape lowers costs and brings greater flexibility with the same reliable performance. When applications on mainframes or legacy systems are more broadly accessible and integrated with other platforms, you can respond faster. That might be to changes needed in the business, to capitalize on market opportunities or simply to give customers what they need.
And with the right combination of tools and platforms, you can keep the investments you've made over the years intact and be a digital, truly connected organization. This range of tools include things like APIs to connect cloud platforms to your legacy data and user interface design that makes applications available on mobile devices with web terminal emulation.
How to get on the path and move fast
Saying this is fine in theory, but in practice investments in mainframe technology haven't exactly been high on businesses agendas — however at odds that might seem with the value of the data they hold.
Given the high value of what's in these legacy systems, any business looking to become data driven simply cannot do that without integrating that information into their core processes. Modifying legacy code is laborious and any number of small mistakes could significantly disrupt the process. Ripping and replacing not only undermines years of investment, but carries its own challenges and risks in data migration. So, what can be done?
APIs hold the key to bringing legacy to life. And the best part of using APIs to connect legacy with the digital enterprise is that the mainframe can be treated the same as all the other assets in your organization that require integration. There are three fundamental considerations for setting it free and getting on the fast track to modern infrastructure:
1 — Take control of your communication standards
Data access or integration is not a new challenge: programmers always found a way to get data in and out of the mainframe. New applications built on Linux, UNIX and Windows were reached with FTP, .NET and Java interfaces. In the 1990s, packaged ERP applications pulled common functions off the mainframe.
However, point-to-point integration becomes overwhelmed by the volume of the new demands and diversity of integration points. Plus, you can't be re-writing integrations every time a new tech standard comes along. APIs are independent of standards and reusable and provide the integration needed. In the past things like RPC, SOAP and REST forced a lot of work to adapt, but with APIs, you can use a standard of your choice and still simplify application interoperability.
2 — Re-use and recycle, especially business logic
Allow developers to call procedures from a mainframe, so they can use those same procedures as building blocks for new applications. This helps you to increase productivity and give your customers new ways to connect for better services.
APIs let you expose the business logic and data of applications as reusable services — build it once and use it over and over again. With the same approach you can extend the capabilities of your legacy systems by consuming services provided by others — internal or external to your organization. This means you can add capabilities or information without having to code it on your own. This is a key factor in removing the risk and expense of replacing or redeveloping your core applications.
3 — Make legacy as easy to work with as ‘new tech'
Convert green screens into user-friendly web-based interfaces and redesign the user experience by improving how one navigates the screen workflow. For instance a .NET wrapper for mainframes can make it more accessible to younger developers with no mainframe experience to start developing new applications.
The best part of using APIs to connect legacy with the digital enterprise is that the mainframe can be treated the same as all the other assets in your organization that require integration. A common integration layer overlaid with an API governance structure secures, protects and manages accessibility. Not only can data fluidly flow to wherever it is needed, but less in-depth coding experience for specific legacy systems is needed to do so.
Enabling SQL access to legacy data lets developers join data from multiple sources. This means that it can be delivered to modern desktop applications used by business users and data scientists alike. Legacy data can also participate in event streams and big data initiatives, be added to data lakes and data hubs, and be accessible to cloud services and AI efforts.
Digital services could overload legacy back ends if they're not modernized. But these back ends could hold the secret sauce to future success. As that data becomes more crucial, and the people who best know how to make use of it retire, it's never been more crucial to extend your API strategy to bring legacy into your modern infrastructure. And it can be done using the same API stack (gateway and portal) that you already use to API-enable your most modern services in the cloud or on the edge.