How Mainframes Remain Essential in the Digital Age

Mainframes aren’t obsolete; they’re just under appreciated—bordering on neglected. Often outsourced and frequently misunderstood, these robust systems are still the backbone of major industries. According to IBM, the most critical workloads are still running on IBM Z: 8 out of the top 10 banks and 7 out of the top 10 insurance companies rely on mainframe technology. From airline tickets to credit card transactions, these operations rely on the secure, dependable environment of an IBM Z. However, what often goes unmentioned is that many of these systems are now operated by a skeleton crew. This minimal staffing is hardly sufficient for diagnosing, troubleshooting, and fixing all the critical systems that run on this powerful platform—a stark contrast to just five years ago.

Remember the days when a mainframe staff from the client side typically included a team of ten? This team would consist of a storage administrator, a CICS programmer, a database administrator, multiple z/OS systems programmers, a capacity planning expert, a data center operations manager, mainframe network specialists, and sysplex personnel. Those were indeed the golden days. So, what led to this change? The primary factors are the dwindling availability of skilled resources and the cost of maintaining such expertise full-time. Additionally, managed services providers often overstate the monitoring they provide and overestimate their system programmers’ skills, who may only be able to follow step-by-step instructions from a run book.

For the past 25 years, I have supported small to mid-sized mainframe clients, primarily preparing them for upgrades to new IBM Z servers. The process was straightforward—we would meet, develop the hardware design, and decide on the necessary OSAs and FICONs. We’d request a week’s worth of data to perform a complimentary IBM zCP3000 study, which utilizes client SMF data to produce detailed reports showing system resource utilization during the selected period. We usually opted for a peak week, as the golden rule has always been to size for the peaks to ensure the new mainframe could handle the current and anticipated growth. Clients would engage actively, asking questions and comparing our estimates with their internal metrics until we reached a consensus. Debates over whether it’s better to have a small number of large processors or a large number of small ones were common. This sizing method is still in use today, but without the crucial client input and internal metrics, it often reduces to a mere IBM exercise aimed at selling an IBM Z.

And how about the way systems are currently managed? We size for the peaks, but what exactly causes these peaks? Is it a natural workload increase, or are there unidentified issues driving up utilization? From my experience, the support staff often lacks the tools, time, and knowledge to pinpoint these problems.

Is help available? Absolutely. You might have heard of several solutions for optimizing your mainframe ecosystem, but none compare to what we offer at SVA Software with zWorkload Reporter (zWR). Our zWR solution doesn’t require installation on your mainframe, understanding that clients are pressed for time. What distinguishes SVA is our team of experts who analyze detailed, management-friendly reports, identify issues in the operational systems, and recommend fixes. Partnering with SVA means adding a dream team to your roster, even surpassing the teams of 20 years ago.

With zWR and SVA’s team of experts, your mainframe will not only receive the care it deserves but also undergo optimizations that reduce costs—a dream come true for any manager.