My day to day life in DevOps engineering is rooted in experience. Occasionally I can remember the magic setting that just works, but it is far more usual that I am starting from a problem (and tool) I’ve never seen before, so I just have to mentally roll my sleeves up and pay close attention to what I am seeing. (The phrase I use most often with other engineers is “show me” and then the close follow up is “let’s try it.”)
This mindset prioritizes experiment over analysis, and it developed over time as I realized how seldom analyzing DevOps engineering problems really helped. Some critical thought is needed to rule out obviously dumb experiments, but after that there are just too many possibilities, and you really just have to try stuff.
I’ve been finding this mindset creeping into my studio work recently. I keep getting questions about vintage amplifiers, advanced plugins, rare guitars, exotic signal processors, and the answer is nearly always “let’s try it and see how it sounds.” If a $50 guitar fits the track, we’ll take it. If the $1000 plugin doesn’t, we’ll mute it.
Just like in DevOps – the problem might be horrible, but the fix might be as simple as a single setting (e.g. don’t leave Entity Framework logging in TRACE mode).