Atlas can be configured to raise alert notification only when the Atlas cluster already scales up to the max tier allowed. This can be help to avoid notification spams which does not require operation team actions and focus on the critical resource consumption alerts like cpu, memory, disk IO.