I agree that it'd be very weird for any of my colleagues to have touched Kubernetes or the build pipeline (unless that was their entire job). Large companies can afford to let people specialize, but that's not necessarily the case with smaller companies where people need to wear many hats.
It's just pretty apparent that I need to put on more hats to be a competitive candidate, especially when the hat I do wear usually isn't relevant to the companies that are hiring. Large companies also have plenty of teams doing more mainstream work (regular backend dev) even if it doesn't include k8s, so they all still have a leg up on me in this sparse job market.
Of course my problem might disappear when the market recovers and the large companies hire again, but I'm not going to sit around waiting for that.
What if I were a a horse-drawn carriage driver and the car industry was relegating my career into niche theme parks? If I don't want to be in that niche anymore (and I don't!), then logically, I should learn how to drive a car and become a driver.
I am just making the observation that the vast majority of available openings that can even remotely fit me (ie. not any sort of frontend) are backend, with a distant runner up of infra. If that is the reality, then I need to meet it.
It's not that I'm against niches. I like the suggestion in the other post to go into infosec. I consider that a niche too but at least most companies would need it. Very few companies make operating systems. I feel that I shouldn't have gone this direction in the first place and this struggle to find a job is a wakeup call to change.
My post already outlines my willingness to learn/retrain to fit a new reality. I'm asking for advice to make such a transition, not really looking to being told to keep doing what I've done when it clearly isn't a very portable area of expertise.
I did start blogging my side projects, but I hadn't considered digging into other implementations and blogging my findings as well. Interesting idea, will try!
Skipping k8s management makes a ton of sense to me, especially since that should be considered more of a DevOps job anyway. But since you say to do CI/CD first, I assume you mean I should go through the entire flow where I submit commits and have the "CD" deploy to the cloud k8s? Seems like I'd still need to be able to bring up my own local instance to see my code changes during dev, and would need to manage the local k8s.
Security might actually be the way to go long-term! The way I see it, it has to be verifiable (not possible to depend on AI) and can't be outsourced (at least not to any adversarial country). I'll definitely look into that!
What was your playbook to get into infosec?
I am not that great at C++, but I guess self-teaching C++ is a more fitting path to go down vs. self-teaching backend tech?
Background:I've had many years of experience, but essentially only 1 year of experience in any one thing. It's all been pretty random, though notably very little frontend. To get specific without being too specific, in recent years I've implemented a filesystem, made a testing framework, added functionality to a C++ library, created a middleman between kernel space and userspace, etc. The...