Question

Difficulty: MediumAuto Scaling and Elasticity

An Auto Scaling group (ASG) uses a step scaling policy based on the `CPUUtilization` metric. The EC2 instances in the ASG require approximately 66 minutes to run bootstrapping scripts and start processing tasks. During traffic spikes, the SysOps administrator observes that the ASG launches additional instances every 55 minutes before the previously launched instances can finish bootstrapping, resulting in over-provisioning. Additionally, the ASG is slow to trigger scaling actions when traffic initially spikes because the CPU utilization data points are evaluated at 55-minute intervals. Which actions should the SysOps administrator take to resolve these issues? (Select TWO.)

  1. Enable detailed monitoring in the launch template used by the Auto Scaling group.Answer
  2. Increase the estimated instance warmup time in the step scaling policy to 360360 seconds.Answer
  3. C
    Increase the default cooldown period of the Auto Scaling group to 360360 seconds.
  4. D
    Configure standard monitoring and reduce the CloudWatch alarm evaluation period to 11 minute.
  5. E
    Attach an IAM policy with `iam:PassRole` permissions to the EC2 instances' instance profile.

Answer

Enable detailed monitoring in the launch template used by the Auto Scaling group, and increase the estimated instance warmup time in the step scaling policy to 360360 seconds.
The correct solution involves enabling detailed monitoring to get 11-minute EC2 metrics, which allows the CloudWatch alarms to react faster, and increasing the estimated instance warmup time to 360360 seconds (66 minutes) so that the step scaling policy allows newly launched instances to complete bootstrapping before initiating further scaling actions.

Step-by-Step Solution

1
Analyze the bootstrapping issue.
The instances require 66 minutes to bootstrap, but the ASG launches additional instances every 55 minutes.
Since a step scaling policy is used, the estimated instance warmup must be adjusted to match or exceed the bootstrapping time (360360 seconds) to prevent the policy from launching more instances before the current ones are ready.
2
Analyze the metric latency issue.
Scaling is slow because CPU utilization data points are evaluated at 55-minute intervals.
To resolve this, detailed monitoring must be enabled in the launch template, which changes the metric reporting interval for CPUUtilization from 55 minutes (standard monitoring) to 11 minute.

Key Concept

Auto Scaling step scaling policies utilize estimated instance warmup to manage scaling frequency during instance bootstrapping, and detailed monitoring enables 11-minute metric granularity to improve scaling responsiveness.
Rate this question