Evangelizing Mainframe
Print Email

The Mainframe–2029

There have been a number of trends in the past few years that make it fairly clear what IBM is thinking in terms of the future direction of the mainframe. So, I thought it would be interesting to extrapolate them and imagine what the world would be like in 2029.
 
Let’s start with the 2029 headlines on IT-related websites. Most are running with a recent prediction that mainframes are doomed and that everyone will be better off moving to the latest trend in computing. Further down in the news is IBM’s announcement of Qz/OS—the latest version of its z/OS operating system that runs on its newly-announced fully-functioning quantum computing hardware. 
 
As everyone knows in 2029, quantum computing uses qubits, which are analogous to bits in a standard computer. Qubits can be in a one or zero quantum state. But they can also be in a superposition of the one and zero states. The interesting thing is that a qubit can be set to zero and one simultaneously, for example, by encoding it in an ion that can spin down for zero, up for one or both ways at once. Qubits enable the machines to process many inputs simultaneously instead of one at a time. They can also be used for artificial intelligence (AI), molecular modelling, financial modelling, weather forecasting and particle physics—anything that requires difficult mathematical problems to be solved, such as decrypting most forms of computer cryptography! A lot of the work came out of IBM’s Yorktown Heights research center in New York and work done by African universities.
 
In 2029, a lot of the mainframe people still tell tall tales about the colorful characters that were around in IT in the early days of mainframe computing, and how large financial companies survived the period of the mainframe veterans retiring and taking all their expertise with them. And how mainframe control was split between those early AI systems, which, looking back, were laughably little better than telling Alexa to play your favorite song. How AI technology has improved in the past ten years—particularly more recently with the introduction of quantum computing hardware. And how Watson Machine Learning for z/OS (WMLz) went from making transactions work better to controlling the whole hybrid sysplex more efficiently than ever before.
 
The other great leap forward around that time was the way the mainframe was opened up to people with no background in mainframes—people who came from the open source community. There had been a project—Zowe—which was the first open-source framework for z/OS. It was announced by the Open Mainframe Project, a collaborative project managed by The Linux Foundation back in 2018. It provided a way for development and operations teams to securely manage, control, script and develop on the mainframe like any other cloud platform. Suddenly, developers didn’t need to have previous mainframe experience! 
 
By using Zowe, non-mainframer developers can use the open-source industry-standard tools they are already familiar with to access mainframe resources and services. They could perform tasks such as administrative, development, test, and operations on z/OS. All of Zowe’s code is open source. It seems strange now to think of the mainframe as being so proprietary even ten years ago. 
 
The other thing, that in 2029 people forget was such a great leap forward, is the ability to run pretty much any operating system under Qz/OS. It all started with IBM announcing z/OS Container Extensions (zCX), which in those days provided a way to run Linux capabilities on z/OS using a Docker container. A Docker container image is a lightweight, standalone, executable package of software that includes everything needed to run an application—code, runtime, system tools, system libraries and settings. Container images become containers when they run on a Docker Engine. The advantage of working this way is that containerized software will always run the same, regardless of the infrastructure. 
 
Containers isolate software from its environment and ensure that it works uniformly. And Docker packaging reduces the complexity and installation of software. zCX runs in a started task and contains both a full Linux system and Docker (a portable software packaging and execution environment). zCX is a virtual appliance that requires minimal involvement after set up. And any management the zCX server needed was simplified by using z/OSMF workflows. So, all of the great features of z/OS (e.g. high-speed networking as well as resiliency around data replication and automation) became available to Linux applications.
 
And that was only the start. Power Systems applications soon became eligible to run under z/OS in the same way as Linux. z/VSE was made available. And we probably all remember the day when it was announced that Windows server could be run in a Docker container on mainframes. Since then, a number of other less frequently-used operating systems have been included to run under z/OS and now Qz/OS.
 
Lots of people still think about mainframes as being enormous. You still get pictures in magazines of mainframes looking as big as three or four double-sized wardrobes. Those days are long gone. It’s been possible to have a normal rack-sized server mainframe for years now. And, obviously, the new quantum computing hardware is still the default size. Although, I’m not sure what they put in all of the available free space.
 
Although IBM has said that there is a pricing premium with the new quantum hardware, their predictions are that larger companies will want the new hardware in order to stay ahead of any security issues around using the older hardware in a world where quantum computing is available. In addition, IBM has announced a new pricing algorithm to make the cost of using Qz/OS on their hardware more affordable for new and existing workloads.
 
So, what do we see in 2029, a mainframe using the latest technology, that’s accessible using open-source software, that can run other operating systems in containers, it’s secure, fast and resilient, and there’s a new pricing algorithm to make the cost of running work cheaper.
 
Doesn’t sound that different to 2019! 
 

Posted: 8/15/2019 11:45:04 AM by Trevor Eddolls

Print Email

Join Now!