I am curious to understand the load capacity of kube-apiserver.
As in if I make a large K8-cluster of let’s just say 100 worker nodes of 100 CPUs each and executing performance runs on it, wherein the pods rely on kube-apiserver a lot.
How can I determine at what point of time the kube-apiserver will crash?
Is there any pre-defined limit on the transactions per second that can be handled by kube-apiserver and if not than on what all parameters does it depends and how to determine the same?