This is an infrastructure improvement that would fix some issues we are having with on-prem installations where the executor layers are killed by autoscaling cluster configurations.
