You’re conflating things. The question was about scaling compute and memory up and down based on load and you’re commenting about never needing to downsize on storage.