AWS Certified Solutions Architect – Associate (SAA-C03)Design Cost-Optimized ArchitecturesMedium
A startup is building a new application that uses a multi-tier architecture with an Application Load Balancer (ALB), EC2 instances in an Auto Scaling Group, and an Amazon RDS database. They expect highly variable traffic patterns throughout the day and want to ensure their compute resources are cost-optimized. Which strategy should they implement to achieve the best cost efficiency for their EC2 instances?
- AUse a scheduled Auto Scaling policy to scale EC2 instances up and down based on predicted traffic patterns.
- BConfigure an Auto Scaling group with target tracking scaling policies based on CPU utilization or ALB request count.
- CPurchase Reserved Instances for the maximum expected number of EC2 instances and use Spot Instances for any additional burst capacity.
- DProvision enough On-Demand EC2 instances to handle peak load and manually scale down during off-peak hours.
Show answer & explanationAnswer & explanation
Correct answer: B. Configure an Auto Scaling group with target tracking scaling policies based on CPU utilization or ALB request count.
Target tracking scaling policies in an Auto Scaling group automatically adjust the number of EC2 instances to maintain a specific metric (e.g., CPU utilization, ALB request count) at a target value. This ensures that compute resources precisely match demand, leading to optimal cost efficiency for highly variable workloads.
Why the other options are wrong
- A. Scheduled scaling is good for predictable patterns but less optimal for 'highly variable' and unexpected spikes/dips compared to dynamic scaling.
- C. Reserved Instances are for predictable, steady-state loads, not highly variable traffic. Spot Instances are for interruptible workloads, which might not be suitable for core application servers in a multi-tier setup without careful design.
- D. Provisioning for peak load with manual scaling is inefficient and costly, as instances run idle during off-peak times.
EC2 Auto Scaling Target Tracking
Target tracking scaling policies for Amazon EC2 Auto Scaling allow you to choose a metric and set a target value. Auto Scaling automatically adjusts the group size to maintain that target.
- Dynamically adjusts EC2 capacity based on actual demand.
- Ensures optimal resource utilization and cost efficiency.
- Common metrics: CPU Utilization, ALB Request Count Per Target.
- Simpler to configure than step scaling policies.
Memory trick: Auto Scaling is like a smart thermostat for your servers: only turn on what you need.