Comment by wmf

Does any container runtime/orchestrator perform this optimization yet? Why wait?