When the load on a system gets close to the capacity of that system - you will typically have to address that situation by increasing the capacity of that system.
You can increase the capacity of a system in two ways: Via scaling - or via performance optimization.
In this video I discuss the different implications for your system and its operational cost depending how you decide to address a capacity issue (via scaling or via performance optimization).
X (Twitter): @jjenkov
LinkedIn: / jakob-jenkov-4a3a8