The mainframe is a beautiful piece of technology, able to securely manage and meet the performance goals of disparate workloads for thousands of users in a relatively small, energy-efficient package. It maintains levels of compatibility and continuity that protect customers' investments in years of development and billions of lines of code. Despite rumors to the contrary, big iron continues to be the back-end system of record for large shops across a spectrum of industries. So, if technology isn't the problem, what's inhibiting the mainframe?
Bumps in the mainframe road
Every year, there seem to be fewer and fewer third-party software products specifically for the mainframe. Most off-the-shelf applications tend to be at a department level and favor smaller machines. As for system and utility software, the big fish continue to eat the small fish and popular tools are concentrated in fewer vendors' hands. This shrinking developer community tends to deflate the mainframe's ecosphere and limit customer choices.
Another issue casting doubt on the mainframe's future is the graying of technical support. The typical mainframe shop has more people who are inching closer to retirement. As the oldsters go, they will take their skills and tribal knowledge of systems and applications with them. This alone may encourage some companies to drop the mainframe rather than risk depending on systems no one can operate.
A corollary to retiring technical support is the loss of mainframe skills. When I graduated in 1981, my fellow alumni knew how to program in PL/1 and COBOL. A few of us even knew our way around JCL. That isn't true anymore, as our universities crank out Java programmers familiar with Linux and Windows. Mainframe shops must spend time and money to train graduates before they are useful. Even worse, some highly technical skills, such as assembler coding and dump reading, get left out altogether, making the customer even more dependent on vendors for technical support.
Lastly, in my mind, the biggest threat to the mainframe's future is cost. Fairly or unfairly, the price of mainframe components makes it an easy target at budget meetings and the first victim when an IT department wants to cut expenses. The alternative -- managing expenses through resource throttling and careful systems management -- has drawbacks, as it leads to wasting hours watching CPU consumption instead of making enhancements.
Controlling costs also leads customers to build asymmetric configurations that are harder to maintain, more likely to fail and no longer aligned with IBM's best practices. IBM has several explanations for why the mainframe may "seem" more expensive, but when shareholders are restless and the CEO wants to cut expenses, "value" becomes less important.
And IBM isn't alone in this. Independent Software Vendors (ISVs) also drive expenses, with contracts based on processor capacity instead of actual usage. These types of contracts tend to balloon customers' software costs any time they upgrade.
The road to a brighter future?
The mainframe's shrinking ecosystem may be irreversible. It most cases, it makes sense to put departmental applications on distributed boxes. The smaller vendor base is explainable by the age of the platform and the fact that the operating system fills many of the ISV software gaps. The good news is that IBM offers z Personal Development Tool (zPDT), which can run a mainframe image on a laptop and is especially valuable to vendors who can't afford to rent or own their own mainframe.
The aging of technical support is surmountable. IBM continues to reach out to colleges to provide mainframe training for interested students. Customers can ensure college hires have a chance to run the big machine as part of their introduction to IT. More importantly, these companies must also stress that the mainframe is still an important infrastructure, not something that will be retired at the first opportunity. Enterprises must also provide viable mainframe career paths that offer growth opportunities to anyone who sticks with it.
That leaves us with cost. IBM shows few signs of relinquishing its grip on the platform and remains the sole provider of processors and the major systems that make a mainframe worth running. While this is good for IBM's revenue stream now, it can make customers wary of entering or expanding the platform. In the long run, this reluctance will ultimately damage IBM's bottom line. Will IBM be able to make the right call when the time comes?
Mapping the right mainframe path
There are still a lot of things to look forward to despite these obstacles. IBM continues to invest in the platform; every generation of its mainframe processor gets faster, more compact, more fault tolerant and more hospitable for non-traditional mainframe workloads, such as Java. The operating system, z/OS, gets more sophisticated and easier to maintain with every release. The transaction processors and database management systems also remain on the cutting-edge of technology, especially as IBM finds ways to make customers' investments play well with modern systems.
The mainframe may not be the only game in town, but I'm convinced it's still one of the best and worth using in the future.
The opinions expressed in the above column belong solely to Robert Crawford and do not reflect those of his employer.
About the author:
Robert Crawford has been a systems programmer for 29 years. While specializing in CICS technical support, he has also worked with VSAM, DB2, IMS and other mainframe products. He has programmed in Assembler, Rexx, C, C++, PL/1 and COBOL. The latest phase in his career is as an operations architect responsible for establishing mainframe strategy and direction for a large insurance company. He works in south Texas, where he lives with his family.