CompTIA Cloud+ (CV0-004)DeploymentHard
A cloud administrator is configuring an auto-scaling group for a web application. The application experiences predictable spikes in traffic during business hours and sudden, unpredictable surges during marketing campaigns. To optimize costs and performance, the administrator needs to define rules that automatically adjust the number of instances based on CPU utilization and also pre-scale the environment during known peak times. Which two types of scaling policies should be implemented?
- AManual Scaling and Dynamic Scaling
- BSimple Scaling and Step Scaling
- CTarget Tracking Scaling and Scheduled Scaling
- DPredictive Scaling and Cooldown Periods
Show answer & explanationAnswer & explanation
Correct answer: C. Target Tracking Scaling and Scheduled Scaling
Target Tracking Scaling allows defining a target metric (e.g., 50% CPU utilization) and automatically adjusts instance count to maintain that target, handling unpredictable surges. Scheduled Scaling allows defining scaling actions at specific times, ideal for predictable spikes during business hours or known campaigns.
Why the other options are wrong
- A. Manual scaling requires human intervention; Dynamic scaling is too broad. This option doesn't specify concrete policies.
- B. Simple and Step scaling are older dynamic scaling types; Target Tracking is generally preferred for metric-based scaling.
- D. Predictive scaling is an advanced feature that uses machine learning for future demand, but typically combined with other policies. Cooldown periods prevent rapid scaling fluctuations, not a scaling type itself.
Target Tracking & Scheduled Scaling
Two distinct auto-scaling policies: Target Tracking maintains a desired metric level, while Scheduled Scaling adjusts capacity at predefined times.
- Target Tracking: Reactive to real-time metrics (e.g., CPU, network).
- Scheduled Scaling: Proactive for predictable load changes.
- Combined for optimal cost and performance in varied workloads.
Memory trick: Track targets, schedule peaks.