A Lack Of Expertise Does Not Force You To Abandon The Mainframe

“Quick! Abandon the mainframe.
…Oh, and by the way, buy a bunch of our platforms and services for millions of dollars!”
—Distributed systems manufacturers

Craig Mullins did quite an insightful piece a while back about the state of things in the mainframe world vis-a-vis the dwindling numbers in the workforce with classic mainframe expertise: COBOL, JCL, CICS, BAL, etc. Many IT organizations in banking, insurance, financial services and retail are taking a cue from this state of affairs, and making plans to abandon the mainframe.

The mainframe—the one platform perfectly suited to the workloads of banking, insurance, financial services and retail—is the platform that these industries rely on more and more as new mobile, web and even IoT workloads come on line.

The point I want to make is that this concern that IT planners have—that they’re going to run out of expertise—and their plans to abandon the mainframe is an over-reaction to FUD. Are we running short on this expertise? Yes. Is it inevitable that we all must abandon the mainframe because of it? No! Are there any other courses of action? Of course there are.

It’s no secret that mainframe systems are selling very well in developing economies and emerging markets, and that the operating system of choice in those markets is zLinux as much as it is z/OS. It started more than years ago, and IBM continues to develop the mainframe/Linux relationship to this day on their newest mainframe, the z14 system.

With Linux or zLinux as the operating system,  planners can leverage any and all of their technical human resources with modern tool expertise (Linux, JavaScript/jQuery, Java, Ajax, WebSphere, WebLogic, XML/SOAP, SQL, etc.) to do productive work on their mainframe platform. They no longer have to rely on the dwindling supply of legacy technology expertise. New applications for the mainframe can be built from the ground up using modern technology developers.

But that’s not your only option. Undoubtedly, if you’re an IT organization leader with significant mainframe assets you have hardware, software, databases and, most importantly, high-value, mission-critical COBOL (or assembler) applications that represent many years and many millions of dollars of development and investment. While one couldn’t be blamed for thinking that the end is near for this tremendous investment; that simply is not the case. No matter what certain chip and distributed platform manufacturers want you to think.

The truth is that your existing assets can be leveraged for as long as you need them. New distributed-based integration solutions can actually leverage your current legacy mainframe assets as they are now. These solutions will allow you to leverage your new generation of technology developers, and let them work with the tools and platforms to which they are accustomed. The solutions provide APIs that allow new applications to interact directly with your legacy applications. You get the best of both worlds: retaining and benefiting from your existing very powerful mainframe assets, while doing more with your new generation of tech developers.

Compare that with the alternative: a complete forklift replacement coupled with completely reengineering a decade’s worth of mainframe-based development. Essentially reinventing the wheel—an extremely expensive wheel. Why should you do that if you don’t have to? And by the way, you don’t have to.

What about new development? Well, these new tools not only leverage existing assets, but also can be used to augment existing assets with new development either on the mainframe or on your distributed systems. Again, you’ll get the best of both worlds: the power and throughput of your existing mainframe systems, and the ability to augment it with new technologies using your new generation of technology developers.

And there are still other options. You can integrate existing mainframe z/OS applications with new applications running on other platforms—any platform, whether distributed x86 or PowerSystems running Linux or UNIX, private or public cloud systems, zLinux mainframe systems or even Watson or mobile. By leveraging near-real-time data replication, legacy databases are synced with new databases, regardless of platform, operating system, database type, or data format, effectively integrating your new apps with your legacy apps.

This option is perfect for mergers and acquisition scenarios where one organization’s IT infrastructure differs significantly with that of the other IT organization. Using this solution, the databases of the two organizations are in sync, while the separate applications run in parallel. Eventually, the best applications can be retained while the other is discarded or archived.

The solution also works well within a single organization, where multiple platforms are run within the data center. The databases for the various platforms can be kept in sync, and the applications used on the less favored platform(s) can eventually be discarded or archived, if that makes business sense.

These are just some of the options open to organizations and IT planners worried about what to do with their mainframe infrastructure in the light of dwindling resources. These are specific options offered by IBM, and my company, DataKinetics. There are actually many more options available from several other mainframe software vendors. Don’t listen to those engaging in FUD over this matter—do some research and find the solution that makes the most business sense to you.

Originally Published on LinkedIn Pulse.