I've built a bunch of simple scripts for this that you can find here [0]. It's not polished for public consumption or updated to k8s 1.10 yet, but it's what I use for production clusters, some of them single-node. Run the preparation script, then the master script, setup an .env file beforehand with the few required variables, and you're good. Feel free to ask questions here or in the repo issues section.
EDIT to add: It's assuming Ubuntu LTS as the node's OS, not sure if that fits your use case. Should be possible to adapt this to ContainerLinux or anything else without much trouble.
I haven't worked with GL's Auto DevOps yet, but I think the cluster should have everything necessary to get going with that.
EDIT to add: It's assuming Ubuntu LTS as the node's OS, not sure if that fits your use case. Should be possible to adapt this to ContainerLinux or anything else without much trouble.
I haven't worked with GL's Auto DevOps yet, but I think the cluster should have everything necessary to get going with that.
[0] https://github.com/seeekr/kubeops