They are the worst, been there a few times, even as "the devops" you get sucked into while(true) loops of tech changes and everybody rewrites everything all the

They are the worst, been there a few times, even as "the devops" you get sucked into while(true) loops of tech changes and everybody rewrites everything all the time to be "cloud ready" and nothing ever gets to production and everybody burns out and leave. I just dodged a bullet recently interviewing with an old boss of mine who wanted to use container orchestration for a perfectly working set of 40+ apps. They use docker in VMs as of now already, but they absolutely want kube or swarm, the reason being that "it's too complicated to manage the capacity planning for the VMs". Yeah let's add an extra layer of clustering and complexity for internal apps that will never scale just so that you feel good about having "an orchestrator", because you don't want to take a couple days to look at CPU/RAM usage and resize your infra.