I'm going to setup a basic jenkins infrastructure to deploy to my website. Very basic stuff.
I want some dedicated machines / compute engines for it. So I'm thinking of using a Kubernetes cluster from Linode, Vultr or DigitalOcean.
I'd like a Jenkins Master to be up and running 24/7 and the worker nodes only on demand. The way I see it is that I have to configure Kubernetes to use certain nodes which I have to pay based on spec and uptime. So I would configure an low-spec k8s master (24/7) and a medium-spec k8s worker node (up only on demand), correct? This is how I would setup a k8s cluster if I would build one on dedicated hardware in my basement. Is the approach comparable? Or is there no need to setup nodes in the cloud at all because every build step of every pipeline will be run inside a container?
Can anyone give me a hint on how to start?
