LAB. 2.3 All in this chapter we are doing at cp-no

Hi @rumpelstilz,

You are correct, typically Pods are scheduled on worker nodes by the control plane's kube-scheduler agent. In order to save on infra costs, the lab setup enables scheduling of user applications on the control plane node as well, through the removal of the control-plane taint following the cluster bootstrapping phase.

While there are exceptions, the user no longer decides where the Pods are deployed - this is the job of the kube-scheduler agent. The user submits a request to the kube-apiserver of the control plane (cp), and upon validation it will be executed and the workload deployed based on decisions made by the control plane agents.

Regards,
-Chris