Which solution will meet these requirements?
Create an Amazon CloudWatch alarm to scale up the EC2 instances when CPU utilization exceeds 90%.
Create a recurring scheduled action to scale up the Auto Scaling group before the expected period of peak demand.
Increase the minimum and maximum number of EC2 instances in the Auto Scaling group during the peak demand period.
Configure an Amazon Simple Notification Service (Amazon SNS) notification to send alerts when there are autoscaling:EC2_INSTANCE_LAUNCH events.
Explanations:
Scaling based on CPU utilization alone may not proactively account for anticipated traffic spikes, as it relies on reactive scaling rather than preemptive scaling.
A recurring scheduled action to scale up the Auto Scaling group before expected peak demand allows the system to be prepared in advance, ensuring availability during high traffic periods.
Manually adjusting the minimum and maximum instance counts is a less flexible solution, and it requires regular monitoring and intervention, which is less efficient than scheduling.
Configuring SNS notifications for launch events does not impact scaling itself and only provides alerts, which doesn’t contribute to proactive scaling.
From what I’ve seen, the answer is:
Create a recurring scheduled action to scale up the Auto Scaling group before the expected period of peak demand.