Talk about an epiphany! There I was at one of the sessions I attended at SHARE in Austin, when someone mentioned something that hadn't really occurred to me yet, and it completely shifted my perspective.
Then, later that same week, someone else mentioned something similar - possibly something I'd heard before but had never connected the dots about.
So, later, I mentioned it in a presentation I gave and got general nods of agreement. Apparently, it's been obvious all along, but no one's been saying anything about it. And here's what it is:
You know all those new "distributed" applications in your organization that have been developed in preference to mainframe ones? Many of them are really mainframe applications that just have some distributed processing and distributed front-ends!
In other words, if you were to turn off your mainframe today, many - possibly most - of your critical distributed applications would stop working because they rely on the mainframe for so much of their important data and processing!
The story that triggered this for me was about a shop that had been exclusively developing "all distributed" applications for so long that management didn't see the value of their mainframe. However, one weekend, there was a bit of an issue with an upgrade on the mainframe that had to be backed out, so it was unavailable for an hour or two during business hours, and suddenly, critical distributed applications stopped working!
It turned out that they were more reliant on the mainframe than ever!
Check it out in your shop and see if it's not the same case: how many of your most critical "distributed" applications have no mainframe interaction at all? Now, how many are so reliant on the mainframe that they'd cease to function if the mainframe were unavailable?
Of course, the reason why no one seems to think of this is because the squeaky wheel gets the grease. I mean, the rarest part of that story is that the mainframe was unavailable - that almost never happens! As a result, nobody notices that their distributed applications are completely reliant on the mainframe... because it's completely reliable!
For me as a mainframer, that's exciting. But for me as a responsible business person, that's scary, because it means that IT management around the world may be uninformed about the fact that the mainframe is still the goose that lays their golden eggs, so they may be reducing their investment in it and even thinking they can move off of it, when in reality, if they did, their goose would be cooked!
Call to action time: what are you doing to make sure your management is aware of how important your mainframe is to the ongoing survival of your business? Since the mainframe is the opposite of a squeaky wheel, someone needs to squawk on its behalf in order to pre-empt some seriously bad decisions!