Question #1294
A company hosts a web application on Amazon EC2 instances behind an Application Load Balancer. The instances are managed by an EC2 Auto Scaling group distributed across multiple Availability Zones. The Auto Scaling group scales up to 25 instances during peak traffic periods but scales down to 4 instances during low-traffic hours. Users are experiencing high latency during sudden traffic surges, though the application stabilizes after a short period.
How should the scaling policy be modified to mitigate latency during traffic surges while maintaining cost efficiency?
Implement a scheduled action to increase the desired capacity to 25 just before peak traffic begins.
Configure a step scaling policy activated at a higher CPU utilization threshold and reduce the cooldown period.
Configure a target tracking policy with a lower CPU utilization threshold and reduce the cooldown period.
Implement a scheduled action to set the minimum and maximum capacity to 25 before peak traffic starts.
Explanation
The high latency during traffic surges occurs because the current scaling policy reacts too slowly. Option C addresses this by:
1. Lower CPU Utilization Threshold: Triggers scaling actions earlier when CPU usage is lower, ensuring instances are added before latency spikes.
2. Reduced Cooldown Period: Allows consecutive scaling actions to occur faster, minimizing delays between scaling events.
Other options are incorrect because:
- A & D: Scheduled actions assume predictable traffic patterns, which may not align with sudden surges. They also risk over-provisioning (cost inefficiency).
- B: A higher CPU threshold delays scaling, worsening latency.
Key Points: Target tracking policies adjust capacity dynamically based on metrics. Lower thresholds and shorter cooldowns improve responsiveness to unpredictable traffic while maintaining cost efficiency.
Answer
The correct answer is: C