From what I understand, mainframes and similar systems have now reached a point where they're quite "sticky". That is, if they haven't already been replaced by

From what I understand, mainframes and similar systems have now reached a point where they're quite "sticky". That is, if they haven't already been replaced by now then there are excellent reasons why they're perhaps going to stick around indefinitely. For one, every time another big ransomware attack occurs, which has cost several big companies months of work and hundreds of millions of dollars to try and clean up, the Big Iron folks look at that and go "Thank goodness we're [relatively] immune to such attacks!" And I once had to personally step in and temporarily revive a zombie but somewhat critical Wintel server which had been attacked in some fashion (the server folks had just given up on it, although finding and correcting the problem itself was pretty trivial), and while I was doing that I remember thinking that such an attack was basically impossible on the bigger systems which I often work on. And so on. Plus, the economics of "scale out" vs. "scale up" don't necessarily look so hot, at least once you reach a certain number of servers. These days you can avoid that to a certain extent my moving at least some of your work to the cloud, but the cloud folks themselves aren't immune to this, either, and I believe they're already starting to figure that out. As far as tech being "an amazing, growing, evolving field", as an old-timer I've grown quite accustomed to looking into some "new" tech only to realize that it's really a rehash of tech from maybe decades ago, just with a new name and a new coat of paint. Given that supposedly at any given moment something like half of the folks in IT have less than five years experience (they're noobs, in order words), I guess that it shouldn't be too surprising that they keep running into problems and reinventing solutions to those problems which were already worked out decades ago.