AWS Certified SysOps Administrator – AssociateCost and Performance OptimizationMedium
A company is running an application on Amazon EC2 instances within an Auto Scaling group. The application experiences predictable spikes in traffic every weekday morning from 9 AM to 11 AM, requiring additional capacity. The current Auto Scaling policy uses target tracking based on CPU utilization, which often reacts too slowly to these sudden spikes, leading to performance degradation. The company wants to ensure sufficient capacity is available proactively and cost-effectively during these peak hours without overprovisioning for the entire day. Which Auto Scaling scaling policy should a SysOps Administrator implement?
- AScheduled scaling policy to increase desired capacity before peak hours.
- BTarget tracking scaling policy based on a custom metric for application latency.
- CStep scaling policy based on network I/O during peak hours.
- DSimple scaling policy based on average request count per target.
Show answer & explanationAnswer & explanation
Correct answer: A. Scheduled scaling policy to increase desired capacity before peak hours.
Scheduled scaling allows you to set up scaling actions based on a predictable schedule. This is ideal for known traffic patterns, such as daily or weekly spikes, enabling proactive scaling before demand increases and preventing performance issues without overprovisioning.
Why the other options are wrong
- B. Target tracking scaling is also reactive, aiming to maintain a specific metric target. While useful, it might still react too slowly to sudden, predictable spikes, and a custom latency metric would still be reactive.
- C. Step scaling reacts to alarms, which might still be too slow for predictable spikes and doesn't proactively address the issue.
- D. Simple scaling, like step scaling, is reactive and will not proactively address predictable spikes in traffic, leading to potential performance degradation.
Scheduled Scaling
A type of Auto Scaling policy that adjusts capacity based on a predictable schedule, ideal for known traffic patterns.
- Proactive capacity adjustment.
- Good for daily, weekly, or monthly traffic spikes.
- Helps prevent performance degradation and manage costs.
Memory trick: Auto Scaling policies dance to different tunes: predictable, reactive, or target-driven.