Modernizing government technology should not mean replacing the dependable mainframe powerhouses that process increasingly complex workloads.
In late April, Rep. Will Hurd (R-Texas) reintroduced the Modernizing Government Technology (MGT) Act. Having cleared the House and currently awaiting consideration by the Senate Committee on Homeland Security and Governmental Affairs, the bill calls for updating outdated government technologies and systems with more modern alternatives offering significant cost and efficiency benefits.
At Compuware, we are advocates of this bill. Government IT is rife with examples of antiquated technologies posing significant risks and absorbing inordinate costs. Modernization is essential to a citizen-first approach – one that provides people with the fastest, most convenient access to digital government services.
However, like top execs in other industries, some government IT leaders are susceptible to dangerous conformities in thought that can lead to irrational decisions. One example is the all-encompassing notion that anything “old” should be replaced, and anything new is automatically better – in terms of system performance (speed and reliability), security and cost. This is not always true, and the mission-critical mainframe is one prime example.
The mainframe has long been shrouded in misperceptions, including the following:
Myth #1: The mainframe is legacy
In IT circles, “legacy” is often a pejorative term, referring to technologies that may have paved the way for subsequent standards and platforms but are now obsolete and inefficient. Mainframes are often unfairly lumped in this category. In reality, mainframes remain highly relevant across industries, including government -- with approximately 225 state and local governments continuing to use them. We’ve seen many organizations try to move off the mainframe, only to quickly backpedal once they discover the “open-heart surgery” their applications and infrastructure must undergo, as well as the cost to maintain and/or secure alternative systems -- all with little to no added benefit.
The explosion in mobility is one major reason for the mainframe’s longevity. Most any mobile application that ends in a transaction -- paying a parking ticket, using Uber or renewing a license online, for example -- ultimately touches a mainframe on the back end. In fact, the term “mobile transaction” is a bit misleading because while the transaction may be instigated on a mobile device, more often than not a mainframe is actually processing it.
IBM’s modern z13 mainframe offers 300 percent more memory and 100 percent more bandwidth than more traditional servers, making it the most powerful computing system on earth. Smartphones may feel light in our pockets, but brawny back-end engines are still needed to support the massive computing loads generated by mobile.
Myth #2: The mainframe represents a huge security risk
Mainframes run a variety of programming languages, including older languages such as COBOL and Fortran. Many erroneously believe these older languages are incapable of supporting the newest security protocols, making the mainframe a major security risk. This is simply not true. COBOL and Fortran continue to be updated, and other aspects of the modern mainframe’s superior security are based in its insular design and function.
All of the hardware and software needed to complete mainframe transactions resides on a single machine, unlike a distributed environment where network traffic can be intercepted by an attacker. In addition, mainframes’ front-end processors often handle the task of interfacing with the rest of the world, freeing up the system to do nothing but what it was expressly designed for: executing transactions. These front-end processors also handle many security aspects, effectively isolating the mainframe from attacks.
The highly publicized data breach of the Office of Personnel Management is often blamed on the mainframe. While mainframe files were accessed, the breach was the result of hackers gaining a foothold in the surrounding distributed infrastructure and accessing mainframe credentials that resided there. Contrary to popular belief, the breach was not the result of the mainframe itself not being secure.
Myth #3: Mainframes cost too much to maintain
Politicians and other government officials love to reference the 75/25 statistic -- referring to government IT teams devoting 75 percent of their budgets to operations and maintenance (i.e. running), leaving only 25 percent or less for innovation. This statistic helps fuel the argument that any older technologies, including mainframes, are a drain and need to be replaced.
Gartner’s most recent IT Key Metrics Data tell a different story. In this analysis of 22 different industries, government (both federal and state and local) fell right in the middle of the pack with regards to percentage of IT budgets allocated to running, growing and transforming IT. This makes it unfair to criticize government IT managers as a particularly profligate and uneconomical lot. It also weakens the implicit underlying correlation -- that heavy reliance on older technology translates to government’s excessive waste.
Technology industry leaders are collaborating to bring greater cost efficiency to mainframe workloads, helping organizations accommodate significant growth in computing loads while eliminating unnecessary overconsumption of resources and keeping costs in check. In addition, studies have shown that mainframe-dependent organizations actually prove to be more cost efficient in the long run than organizations using large, distributed, commodity server-based infrastructures.
Myth #4: Mainframe talent is quickly evaporating
It’s true that the largest mainframe talent pool, the baby boomers, are retiring. But this doesn’t mean it’s time to rip and replace these systems. Rather, mainframe environments simply need to be modernized, making it easier for any developer to work with COBOL -- a language easy for developers to add to their programming repertoire. Examples include replacing the antiquated green-screen development environment with a modern, familiar integrated development environment, as well as leveraging Java-like tools and technologies that make updating and maintaining COBOL applications easier and less risky for newer developers.
Mobile applications are often multiplatform in nature, spanning several systems before ultimately landing on a mainframe. Developers working on these applications increasingly must be able to toggle seamlessly across them. Providing government technologists with a multiplatform DevOps environment will enable them to quickly and accurately complete modifications, truly understand the mechanics of how their systems work and ultimately reduce if not eliminate technical debt. In fact, removing unnecessary obstacles around the mainframe will likely get developers excited about the prospect of working on some of the government’s most critical digital initiatives.
Myth #5: Choosing to stick with the mainframe means we forgo the benefits of the cloud
The cloud offers many benefits, including instant scalability, cost-efficiency and flexibility. Contrary to popular belief, the decision to stay on the mainframe does not mean an organization has to forgo these benefits. Instead, there’s an opportunity to combine the greatest attributes of both platforms. The key is understanding which types of applications are best suited to each environment. Generally, any mission-critical application serving as the lifeblood of an enterprise should stay on-premise on the mainframe, while other less critical applications like human resources and email can benefit from the cost efficiencies of shared cloud resources.
Imagine never having to enter a motor vehicles department to renew a license ever again. Services like this are at the heart of the citizen-first vision, and modernization through the MGT Act and other initiatives will indeed be critical to getting there. But a citizen-first mission also means being the best possible stewards of taxpayer resources, which means we must not apply blanket generalizations to all technologies.
The mainframe is one prime example, representing huge investments and whose unrivaled security, reliability and computing power can – and should – continue to be harnessed. Just because the mainframe is old, does not mean it has outlived its utility. With better understanding and a fair dose of modernization, the mainframe will fulfill a unique, indispensable role in citizen-first IT service delivery.