What’s wrong with Ansible?

Tribalism will affect how this is received, like cursing out vi or apple in a crowded room, but it’s important to see what else is out there and what they offer. Hint: If Ansible is bolting things onto the side of itself like event-driven triggers and connecting to AWX, then you have a good idea of what Ansible needs crutches to do and keep up to last-gen tech. One can only bolt so many bags on the side before the entirety falls apart, and IBM no longer has the goodwill to keep enthusiasts doing the heavy-lifting – even if IBM is repeating what Canonical did a decade or more ago without repercussion.

Patching shouldn’t need an automation scaffolding. I’ll leave that there, that it’s entirely possible to patch your systems in a very automated, patchset-promoted fashion and not need to touch what we currently call Automation. I’ve seen and done it 20+ years, but to be fair that’s only how long I’ve been in the Enterprise space where that was the focus vs the relaxed tolerances of the soho/robo market.

This-gen tech is responsive and self-organizing from the ground up, and responds in real-time to changes. Comically, it’s usually a collection of well-established components like consul that powers the this-gen stuff.

I joined a job with this holy trinity, but they pay the tax every paycheque. I “dead sea” left a toxic mess with failing puppet managers a FIN coup had installed but with good tooling, to a great environment with known faces and good management left behind after their arrogant toxicity couldn’t cope with remote-first workers and bailed. The fact the tooling is complete shite is just a feature we cope with in this awesome environment, and while the environment stays excellent we’ll solve that technical challenge or we’ll bail if the environment gets toxic again first.