The Future for Enterprise Optimization is a Mainframe and Cloud Strategy
Mainframe workloads continue to grow in both magnitude and business value. As a matter of fact, the platform destined for extinction isn’t the mainframe, as predicted, but rather on-premises distributed infrastructure. That’s why the optimal future architecture for large enterprises includes only two platforms: the mainframe and the cloud.
Those seeking to optimize the performance, reliability, adaptability, security and IT economics of the digital enterprise should therefore strongly consider a two-platform strategy that dispenses with costly, unreliable and impossible-to-secure on-premises distributed infrastructure.
What to Eliminate
IT is under intense pressure to deliver value digitally in relentlessly expanding ways. It must:
Deliver compelling real-time client experiences
Deliver sophisticated analytics to diverse decision-makers
Empower people both within and beyond the walls of the enterprise to collaborate without friction
For several decades, IT has relied heavily on on-premises distributed infrastructure to achieve such goals. At first, that was with good reason. Innovative client off-the-shelf applications and custom development tools were closely tied to distributed server platforms—rather than the mainframe, which had become a very closed environment. There was as yet no cloud-based Infrastructure as a Service (IaaS), Platform as a Service (PaaS) or Software as a Service (SaaS) offerings available that would allow enterprises to offload the burdens of distributed infrastructure ownership.
Those burdens, however, now vastly exceed anything anyone ever imagined. Enterprises have piled generation after generation of server, storage and networking gear on top of each other. The complexity of these massive pieced-together environments makes them inherently unmanageable. IT spends untold budget dollars on infrastructure management staff—and tools upon tools—yet still struggles to maintain service and security levels.
As these ownership burdens have made on-premises less practical, IT has:
- Thrown more people and technology at the challenges of service level management, security, workload optimization, business continuity, etc.
- Offloaded as many existing and future workloads as possible to the cloud to achieve lower cost, greater flexibility, stronger security, etc.
- Virtualized its infrastructure elements in order to treat them as a single resource pool—which essentially amounts to an attempt to recreate the mainframe
These trends clearly point to an obvious conclusion: It’s probably best for enterprise IT to dispense with the burden of on-premises distributed infrastructure altogether.
The Two-Platform Enterprise
Most IT leaders can’t imagine an enterprise without on-premises distributed infrastructure. That’s because it has become a bad habit to throw good money after bad into a bottomless pit of incomprehensible complexity that offers no competitive advantage.
But enterprises can thrive without on-premises distributed computing. A two-platform environment of mainframe and cloud can work for any organization seeking maximum digital effectiveness.
The mainframe runs the core systems of record that define the business. Those applications have been tightly honed to the needs of the business over decades. And no platform offers a more scalable, reliable, high-performance, secure and efficient total cost of ownership place to run them.
That doesn’t mean mainframe environments are perfect as-is. On the contrary, mainframe development/test and operations processes at most organizations have become ossified—crippling IT’s ability to modify and innovate on the mainframe as nimbly as required in today’s fast-moving markets. Most organizations also face the loss of their most experienced mainframe professionals.
These issues, however, don’t require re-platforming. Instead, they necessitate new tools and processes. Those tools empower non-mainframe DevOps staff to understand, change and improve poorly documented COBOL applications. They simplify unit testing on new mainframe code. And they integrate mainframe DevOps into the multi-OS environment through popular solutions like Jenkins and Git.
The cloud runs everything else. Payroll and email systems don’t provide competitive differentiation, so why treat them as anything but commodities? SaaS is cheaper, more mobile-friendly, and delivers new capabilities immediately and non-disruptively.
And if any business-differentiating applications need to run on VMs, why own the underlying infrastructure? It doesn’t make operational or economic sense to do so now that PaaS/IaaS solutions allow organizations to rent capacity on-demand.
What’s Your Motivation?
The move to two-platform IT can feel extreme. But given the pressures on IT to do more with less—and do it more quickly and with more stability than ever before—two-platform IT is the best decision.
No other strategy gives IT the ability to so dramatically:
- Innovate at speed. Once you achieve Agile and DevOps on the mainframe, you can bring digital innovation to market with unmatched speed. This is possible because you’re modifying your core business logic—whether it resides in your systems of engagement or systems of record—directly as required, rather than constantly wasting time and money on needless re-platforming.
- Optimize service levels at any scale. The combination of mainframe and cloud can reliably deliver outstanding performance at any peak workload without massive capital spending. On-premises distributed infrastructure, on the other hand, locks IT into an interminable losing battle to remove chronic bottlenecks—even if it’s highly virtualized.
- Shift resources from non-productive maintenance to true value delivery. The massive volumes of time, money and energy that on-premises infrastructure consumes just to “keep the lights on” should be channeled into more beneficial projects. Two-platform IT eliminates this resource drain.
- Minimize digital security and compliance risks. Two-platform IT is far easier and less expensive to defend and govern than Rube Goldberg-style distributed infrastructure.
While two-platform IT may, at first glance, seem like a contrarian approach, it’s actually highly rational. It has proven to work well for organizations. And it can work well for any enterprise that runs on the mainframe—and that recognizes the superiority of the cloud to traditional on-premises distributed infrastructure.