Autoscaling adjusts a service's number of instances according to load (requests, CPU, queue), within minimum and maximum bounds. It absorbs spikes and reduces cost when idle; setting the minimum to zero removes the idle cost, but adds a cold start.
Autoscaling adjusts a service's number of instances according to load (requests, CPU, queue), within minimum and maximum bounds. It absorbs spikes and reduces cost when idle; setting the minimum to zero removes the idle cost, but adds a cold start.