Who says that IT has to be complicated?
If there were one, single maxim for the IT industry, it is: "The only constant is change." Surely, anyone who has spent more than a few days in this business could figure that out on their own. However, I have noticed that a lot of the change is due to an inexplicable movement towards increased complexity disguised as innovation. This is -- to put it bluntly -- complete idiocy.
I like to think of myself as some sort of tech-savvy Luddite, as silly as that sounds. Basically, I appreciate technological innovations, but I do not sign onto deploying fancy, new technology just because we can. I am apt to lag behind a bit and stick with stable, proven technology. Why? Because oftentimes, it just works. Moreover, it's good enough.
I attribute my philosophy to the fact that I grew up in the era before the Cloud when we had to build our own data centers, acquire physical computers, rack and cable those computers, and get paged in the middle of the night when the feces hit the flabellum. There were no little AI-powered DevOps bots that would spin up new containers to replace the broken ones. And, we walked ten miles to school...in the snow...uphill...both ways!
Old man comments aside: I wish that more IT practitioners would take a chill pill, and investigate and embrace proven solutions that will work for a significant number of problem domains. One does not always need to reach out for the shiny, new thing; some really smart folks have put a lot of thought into fixing a whole bunch of problems. Plus, those solutions have had lots of soak time and have persevered. Look at the issues you're facing today, find analogous ones from the past, and see how your predecessors worked it out. Perhaps, you might need to make a few tweaks here and there. But, in the long run, that requires less effort and drama than deploying a whole new stack/ecosystem.
I am not naive and do realize that there are organizations out there that have really hard problems like processing metric crap-tons of data in real time. But, most people do not have problems that hard. In my travels, I come across a lot of small and medium data. Traditional infrastructure can easily process that sort of stuff. In other words, use your friendly RDBMS rather than the NoSQL, stream processing, microservice framework that runs within a web browser (or some other equally ridiculous technology).
At the same time, do not be afraid to incorporate newer technologies when it makes sense. Heck, even us Luddites embrace the Cloud to get massive computing power so that we can get to our beers faster...
tl;dr version:
Even shorter version: K.I.S.S.
I like to think of myself as some sort of tech-savvy Luddite, as silly as that sounds. Basically, I appreciate technological innovations, but I do not sign onto deploying fancy, new technology just because we can. I am apt to lag behind a bit and stick with stable, proven technology. Why? Because oftentimes, it just works. Moreover, it's good enough.
I attribute my philosophy to the fact that I grew up in the era before the Cloud when we had to build our own data centers, acquire physical computers, rack and cable those computers, and get paged in the middle of the night when the feces hit the flabellum. There were no little AI-powered DevOps bots that would spin up new containers to replace the broken ones. And, we walked ten miles to school...in the snow...uphill...both ways!
Old man comments aside: I wish that more IT practitioners would take a chill pill, and investigate and embrace proven solutions that will work for a significant number of problem domains. One does not always need to reach out for the shiny, new thing; some really smart folks have put a lot of thought into fixing a whole bunch of problems. Plus, those solutions have had lots of soak time and have persevered. Look at the issues you're facing today, find analogous ones from the past, and see how your predecessors worked it out. Perhaps, you might need to make a few tweaks here and there. But, in the long run, that requires less effort and drama than deploying a whole new stack/ecosystem.
I am not naive and do realize that there are organizations out there that have really hard problems like processing metric crap-tons of data in real time. But, most people do not have problems that hard. In my travels, I come across a lot of small and medium data. Traditional infrastructure can easily process that sort of stuff. In other words, use your friendly RDBMS rather than the NoSQL, stream processing, microservice framework that runs within a web browser (or some other equally ridiculous technology).
At the same time, do not be afraid to incorporate newer technologies when it makes sense. Heck, even us Luddites embrace the Cloud to get massive computing power so that we can get to our beers faster...
tl;dr version:
- Look at your business problems.
- Get as deep an understanding of those problems that you can.
- Find existing, proven solutions that would work to solve those problems (or a chunk of the problem, as the case might be).
- Look at the new-fangled, wiz bang stuff to augment the areas where the proven tech doesn't quite fit.
- Lather, rinse, repeat.
Even shorter version: K.I.S.S.