It is often said that the most difficult hurdle in modernising mainframe environments is the inherent risk involved in making a change to a core business system that has been working well for the past 30 or 40 years. However, there are many challenges associated with opting to maintain applications written in Natural and integrating with Adabas databases.
Exorbitant operating costs
Mainframes are expensive. Exactly how expensive depends on several factors including size, complexity, usage and licensing fees. Based on insights from our customers, a mainframe's annual per MIPS cost falls somewhere between £1,500 and £3,800. A comparable workload can run in a Public Cloud environment for just 8-12 per cent of that. This potential for infrastructure savings became a reality for The New York Times when they realised an astounding 70 per cent cost saving by migrating their CIS / Aristo application group from a mainframe to AWS with Advanced in 2017.
Software AG products like Natural, which commands substantial licensing fees on its own, exacerbate the cost conundrum.
The 2020 Mainframe Modernisation Business Barometer Report, found that organisations could save on average £24 million if they modernised the most urgent aspect of their legacy systems. While individual results can certainly vary, modernisation is a worthy consideration, and armed with the right tools, can be a relatively low-risk endeavour.
Diminishing skills and resources
Developers who understand languages such as Natural and DBAs who know databases like Adabas are incredibly difficult to find, with the mainframe-era workforce steadily retiring. To make matters worse, most universities no longer offer mainframe instruction since no one would dream of using these technologies for greenfield development projects. According to a 2018 Forrester Consulting study, enterprises have lost an average of 23 per cent of specialised mainframe staff in the last five years. A huge 63 per cent of these vacancies remain unfilled. As the shortage of experienced developers and DBAs grows, the risk and expense of relying on a shrinking talent pool will continue to rise.
Sluggish response to competitive pressure
Lack of frameworks, productive and advanced IDEs, debugging tools, and test automation add significant time to development cycles on the mainframe. Organisations relying on maintaining and extending legacy systems using waterfall development methods have a very long average time-to-market for new business needs and respond slowly to challenges from competitors.
In recent years, digital transformation, underpinned by modern agile practices such as continuous integration and DevOps, has further exacerbated the troubles associated with sluggish development cycles in legacy environments. According to the 2020 Mainframe Modernisation Business Barometer Report, 85 per cent of respondents preferred agile development practices over traditional waterfall methods, with 33 per cent stating that modernising would allow them to be more reactive to market changes.
To make matters worse, the cumbersome nature of these old systems has significant downstream effects. Even when new applications are developed using modern technologies, integrating these with core business functionality running on mainframes is a highly constrained, time intensive, and risky task.
Realising the state of agility required is a difficult prospect, especially for businesses tethered down by cumbersome systems. The impact and breadth of these legacy systems is not to be underestimated.
The low-risk, phased modernisation path
Fortunately, the future is not as grim as it may seem. You can retain your Natural code as-is while taking advantage of the cost savings and infinite elasticity of the Cloud by rehosting your legacy estate on our Application Transparency Platform (ATP).
Rehosting, sometimes referred to as replatforming, is ideal for companies whose applications are still meeting their business needs, and who therefore wish to retain their Natural development skills while moving away from non-relational databases and expensive mainframe infrastructure. By rehosting with ATP, your Natural code is migrated onto modern, distributed platforms, either on-premises or in the Cloud, without change. Underlying legacy databases are migrated to relational models, configured to seamlessly interact with your applications via ATP. You can also use ATP to automatically refactor syntax to COBOL, Java, or C# on demand. As an alternative to re-engineering and conversion solutions available for Natural and Adabas, ATP dramatically reduces the timeframe for the replacement cycle and accelerates your return on investment.
Our upcoming webinar, ‘Eliminate mainframe and licensing costs without changing your Natural syntax’ on Wednesday 21 October at 10:00 EDT | 15:00 BST will share insight into how this phased approach to modernisation removes the common risks many executives and application developers cite as reasons to stay on the mainframe. Cindy Howard, Vice President of Technology and Solutions, will discuss why rehosting or refactoring Natural with ATP, and migrating Adabas to relational, is key and how our solutions allow you to modernise at your own pace. Register your place here.