AWS Certified Solutions Architect Professional SAP-C02 Practice Question

The correct answer is to implement a combination of scheduled scaling and step scaling policies. Scheduled scaling is the most cost-effective way to handle predictable traffic patterns, such as scaling up for business hours and scaling down during nights and weekends. This directly addresses the over-provisioning issue during off-peak times. For the brief, intense, and unpredictable spikes, a step scaling policy provides more granular and aggressive control than a target tracking policy. It allows for configuring different scaling responses based on the magnitude of the alarm breach, ensuring that a sudden, intense spike triggers a correspondingly large and rapid increase in capacity, which is crucial for performance.

• Using only a target tracking policy is a plausible but less optimal solution. While it adapts to load changes, it may not react aggressively enough to the described 'intense spikes' compared to a well-configured step policy, and it is less efficient at managing the known business-hour schedule than proactive scheduled scaling.
• Using predictive scaling is incorrect because it relies on forecasting from historical data for cyclical patterns. It is not designed to handle sudden, unpredictable spikes in demand.
• Implementing a custom solution with a Lambda function to modify the Auto Scaling group adds unnecessary operational overhead and complexity. AWS Auto Scaling provides native policies that can achieve the desired outcome more simply and robustly.