Event-based scaling triggers resource adjustments based on predefined events rather than actual compute utilization or system load. In this case, the system is scaling up every time
2,000 new users subscribe, regardless of whether the underlying compute resources are actually under stress. This can lead to unnecessary scaling, increasing costs without improving performance.
To optimize scaling, the engineering team should consider a hybrid approach that combines event-based triggers with resource utilization metrics (e.g., CPU, memory, or network load) to prevent over-provisioning.