Mainframe futures – a staff perspective
Monday, 3 April 2023
Let me put my cards on the table from the very outset, I believe mainframes to be one of the most secure and versatile platforms for computing that there is. There is no ‘but’ or ‘however’ following that sentence. What I would like to say is that other platforms are ideal for certain types of computing. I use some of the many apps on my phone all the time. It’s quick and convenient. I use Word and Excel and the Adobe desktop products on my laptop all the time. And I know that cloud computing is brilliant for certain types of work, data analysis. In any reasonably-sized computing department, all these platforms should be used optimally to get the most out of them.
Unfortunately, IT people seem to be as fashion conscious as the Kardashians! The move away from the mainframe seems to have been going on since the 1990s. People have said, that client/server will replace mainframes, or mid-range systems, eg the Dec Vax line, Tandem, Sun Microsystems, and many other systems including IBM’s AS/400s and RS/6000s. And now we have people saying that cloud computing will completely replace – the word of choice is ‘modernize’ – mainframes out of existence. If history tells us anything, then it’s that that won’t be true!
IBM has continued to develop their mainframes, operating systems, and applications. And they have continued to absorb non-mainframe technologies onto the mainframe. There was a time when all mainframe communication to devices used SNA. But now, and for decades, it uses the same communication methods as everyone else, TCP/IP. Linux runs on a mainframe, in fact, it can be the sole operating system on mainframe hardware. So, all those Unix specialists can use mainframes. There’s a whole lot of other software available that makes the mainframe usable by computing specialists with no mainframe expertise. Ansible can be used to automate so many mainframe processes. Programmers can use Git for application development. z/OS Container Extensions (zCX) makes Docker available on mainframes. IBM’s z/OS Management Facility (z/OSMF) interface is used to provision, deprovision, and otherwise maintain zCX instances.
In fact, IBM z/OS Management Facility (z/OSMF) provides system management functions in a task-oriented, web browser-based user interface with integrated user assistance. This improves z/OS system programmer’s productivity by using simplified, streamlined, and automated tasks.
Visual Studio Code (VSCode) is available. For developers using blockchain, IBM has released a VSCode extension called the IBM Blockchain Platform Extension, which helps developers create, test, and deploy smart contracts, enabling users to connect to a local Hyperledger Fabric (HLF) development environment to test smart contracts before they are deployed in runtimes such as IBM Cloud (with IBM Blockchain Platform for IBM Cloud) etc.
Broadcom has released Code4z, which is a collection of free extensions for Microsoft Visual Studio Code tailored for developers on the mainframe platform.
Zowe provides another way for people without mainframe experience to treat mainframes like any other servers. It’s an open-source software framework that provides easier interoperability and scalability between products and solutions from different vendors. And there is a Zowe Explorer Extension for VSCode, making working on mainframes a lot easier for non-mainframe people.
Lastly, Z Open Automation Utilities (ZOAU) provides a runtime to support the execution of automation tasks on z/OS through Java, Python, and shell commands. With ZOAU, it’s possible to run traditional MVS commands, such as IEBCOPY, IDCAMS, and IKJEFT01, as well as perform a number of data set operations in most scripting languages.
You get the idea – the mainframe is more than just green screens and arcane applications, it can be used by non-mainframe-trained IT specialists.
But that’s the state of play at the start of 2023. This is a world where we know lots of mainframe specialists are planning to retire in the next five years (although, some of them have been saying that for a good few years now!). And we know, although a lot of non-mainframers may not, that there is an easy-ish route into working on the mainframe for non-mainframers, But what else can IBM do to ensure the continued success of mainframes?
I was taken to task the other day because I moaned that universities weren’t training people in COBOL and mainframe skills. It was pointed out to me that colleges take time to change their courses. So, even if a lecturer wanted to add COBOL or mainframe skills to the syllabus, they would have to wait until the next review.
The answer to my question seems to be the inclusion of AI. We saw IBM’s plans for z/OS 3.1 has lots of AI built into it. I’m guessing, but I would expect this to be the start of more-and-more AI appearing in the software. This, I’m expecting, will start to deskill a lot of the jobs currently performed by mainframe specialists.
So, looking at mainframe futures from a staff perspective, we can see that mainframes will continue to be a viable computing platform. We can see that IBM and others have taken lots of steps to ensure that non-mainframe specialists can treat the mainframe as simply another server, and perform work on it with little extra training. Ansible runs from a laptop, for example, but affects the whole mainframe environment. That should allow ageing mainframers to take the retirement that they deserve without leaving organizations without anyone who can work on the mainframe. And, if my prediction is correct, the growing use of AI will deskill a large percentage of mainframe staff, making it easier to employ people to do work that, only a few years ago, would have required large amounts of technical expertise. And that means fewer mainframe experts will be needed by an organization.
Maybe one day, a mainframer will simply need to say, “Alexa, IPL the mainframe”. And it will!
If you need anything written, contact Trevor Eddolls at iTech-Ed.
Telephone number and street address are shown here.