Initial Readiness Delay

The equivalent of the Kubernetes HPA --horizontal-pod-autoscaler-initial-readiness-delay flag can be set by providing the parameter initialReadinessDelay in the predicitve config YAML. Unlike the HPA this does not need to be set as a flag for the kube-controller-manager on the master node, and can be autoscaler specific.

This option is set in seconds.
This option has a default value of 30 (30 seconds).
See the configuration reference for more details.