Business infrastructure has changed dramatically over the last decade, thanks in no small part to the growth in data-centric activities, the rise of artificial intelligence (AI) and machine learning, as well as tighter regulation of data. In spite of these changes – and the belief that mainframes would die out around the turn of the millennium – many companies continue to depend on, and invest in, their mainframe estate.
Digital transformation efforts over recent years have signaled less of a wholesale move to the cloud than many may have predicted. This is especially true in traditional mainframe heartland markets such as financial services and retail. According to the latest Ensono-Cloud Industry Forum survey, just over half (53%) of businesses currently support a mainframe estate. This rose to almost two thirds (64%) among large companies with more than 5,000 employees – hardly a death knell for mainframe in 2018.
How exactly has mainframe withstood the test of time? To some extent, it’s down to reliability; naturally, mainframes still prove to be useful in applications where frequent or prolonged downtime would be costly or catastrophic. There are also stringent regulations to consider, with many companies deciding that the best place for sensitive customer data remains within a modern mainframe environment under their own purview. This is especially true now that the EU’s General Data Protection Regulation (GDPR) has come into play.
In addition, the long-lasting appeal of mainframe can be traced to its adaptability over the years: they’re increasingly used to handle the big data applications that are shaping society. From the seemingly unstoppable internet of things (IoT), as well as the industrial internet of things (IIoT), automation and machine learning is sweeping through the workplace, and the mainframe is often there as the backbone.
The kind of mass data manipulation that comes with these technologies lends itself perfectly to the mainframe versus other server platforms. The increase in transactions hitting the backend as a result of changing user habits and increased use of mobile applications has had a profound impact on the mainframe, highlighting the relevance and importance of the platform in volume transactional data scenarios. People check their bank balances far more often than they used to thanks to mobile devices. Each one of those look-ups is a transactional hit. The same applies for things like pre-loading retail baskets as well as mobile payments and casual messaging. This growth in the kind of data that mainframes were built for is at the core of why the mainframe is thriving in this new mobile-centric business world.
The Negative Stigma
Whatever positive reasons there might be for the endurance of mainframes, there’s undeniably some negatives too. Companies see it as an explicit barrier to digital transformation efforts.
Some fear the repercussions of making tweaks: 46% of respondents to our survey cited that making any changes to legacy systems would mean major business disruption. To use the adage, “if it isn’t broken, don’t fix it”.
For other companies, the main barrier is one of internal resources. One-in-three (30%) admit that they would only have a limited skills base internally who would know how to maintain an updated system. The mainframe space is already struggling with one of the biggest deficits in the IT skills gap, making any wholesale changes or migrations all the more complex to achieve.
Other mitigating factors include the scale of the existing systems, their criticality to the operations of the business, as well as simple inertia within environments.
Reconsidering the Transformational Impact of Mainframe
Updating the mainframe environment so that it becomes part of a business transformation strategy, rather than an obstacle, is the logical outcome. Begin by designing and building APIs that expose existing legacy business logic (services) through restful interfaces as a first step. At this point you can interface with modern platforms such as Linux or Mobile in a way that is consistent with those platforms’ expectations. From there, updating and evolving the COBOL components behind the APIs is the next step, albeit a more time and skills intensive one. Now you’re ready to include the mainframe in modern architectural approaches such as microservices and, in addition to COBOL backed services, you can implement in node.JS, Java or Swift, just as you would on other platforms. With this direction, companies can retain the resiliency, performance/scale and security features associated with the mainframe, while still benefiting from all the advantages of a modern, agile DevOps approach.
Mainframes continue to deliver significant compute performance across many sectors. They operate at the core of many companies to reliably, consistently and securely process data and transactions. They should be embraced as core components, ripe for transformation, rather than viewed as a legacy that must somehow be overcome. To simply disregard them in any transformation journey is to ignore the very DNA of the companies.
Continuing reading: Why the Mainframe Demands Relevance in a Cloud-Centric World