Remember when you got your driver’s license? That was a pretty big day. Now imagine that as the person at the DMV handed you your card, he or she also slipped you permits to operate a boat, drive a motorcycle and fly the Space Shuttle. Sounds crazy, right? Yet it’s happening in the world of mainframes every day thanks to new tools that make it possible to program and manage big iron in just about any language and on any platform.
Even five years ago it would have been inconceivable for this to happen. If you wanted to use a mainframe, you had to know COBOL. That’s all fine and good—if you graduated from college in 1978. But what about the next generation of mainframers? How were they supposed to use computers that required users to know a language that wasn’t even taught in most computer science programs anymore?
+ Also on Network World: IBM buys EZSource to help transform your mainframe apps for the digital world +
This skills gap is one of the biggest challenges organizations that rely on mainframes face, but it’s rapidly evaporating—and not because COBOL is suddenly cool again. Instead, new interfaces are making it possible to program a mainframe and even integrate with COBOL programs without actually knowing COBOL. That means users can use R, PERL, Python, PHP and other languages that are in common use today. All they need is a user ID and password on the mainframe.
What you can do without having to be a mainframe expert
What can users do on mainframes without having to actually be experts in using mainframes? Here are a few examples (all of which are really happening right now):
- Deploy your R statistical models that you built to run with Apache Spark from your workstation/laptop to z/OS.
- As a PHP programmer, you could set up and run WordPress on z/OS for tighter integration with existing z/OS applications.
- As a Python or PERL programmer, you can use the bash shell program and continue to work the same way as if you were running on Linux or macOS.
- As a database administrator, you could command a relational database with 10TB of memory and 168 cores running on a set of 5-GHz oct-core processors! And by the way, this is the largest commercial symmetric multiprocessor in the world that can additionally be clustered with 31 other mainframes as one logical system without partitioning or sharding.
I talk about mainframes a lot, and I’m often asked why anyone would do any of these things on mainframe when there are other platforms that can do the job. It’s a great question, and the answer is why I’m so excited about what mainframes can do—and why you should be, too. Really important data lives on the mainframe, and being able to do real-time, complex analytics against this data—without having to copy the data to another system first—is game-changing. Let that sink in. Imagine being able to do college assignments and work with mission-critical information on one box.
Now imagine being about to do all of that in contemporary languages that you already know without having to learn COBOL. And then imagine being able to use your laptop, tablet or smartphone rather than a green-screen terminal emulator. All of a sudden using a mainframe isn’t so exotic or confusing. It’s just a computer. A really fast one.