Scale-In

Scale-in means shrinking back the number of servers you were running in parallel once you no longer need them all. It's the same idea as opening extra checkout counters when a store gets busy, then closing them again once things quiet down.

Cloud services are often set up to automatically adjust the number of servers based on how many people are using them. During busy periods, more servers are added — that's called scale-out. When traffic quiets down, servers are removed again through scale-in, saving money on capacity that's no longer needed.

This is different from changing the performance of a single server. Scale-in and scale-out always refer to the number of servers, while adjusting a server's own performance is a separate concept.