passdrill
AWS SAA-C03 · EC2 & Compute · Card 017/024 medium

An Auto Scaling group serves an application with a slow, multi-minute boot process. The team configures a warm pool with instances kept in the Stopped state to cut scale-out latency. Immediately after setup, while instances sit in the warm pool but have not yet been drawn into service, how do they affect the group's reported desired capacity?

  1. Each pooled instance immediately counts toward the group's desired capacity, the same as an InService instance
  2. Pooled instances double-count, appearing in both the warm pool metrics and the desired capacity
  3. The warm pool cannot exist unless desired capacity is first reduced to zero
  4. Pooled instances sit outside the desired capacity and only begin counting toward it once they leave the pool and enter service
Next card → Shuffle