SAP-C02 Continuous Improvement for Existing Solutions Practice Question

Correct answer & explanation

Option B is correct because increasing the target CPU utilization threshold and decreasing the cooldown period may cause more aggressive scaling but does not address the 5-minute startup lag. Predictive scaling (C) is proactive but still has the same startup delay. Option D (pre-warming) is not a standard AWS feature. Option A is correct because using a scheduled scaling policy to add instances before known peak hours can pre-empt the startup delay for predictable traffic patterns. However, the question asks for MOST effective; if traffic spikes are unpredictable, using a larger instance type (E) would reduce per-instance CPU but increase cost. The best answer is to reduce the cooldown period and set a lower CPU threshold (B) but that still has lag. Actually, the best is to use a weighted target group with EC2 instances in a warm pool that are already registered with the ALB but not serving traffic until needed. Since warm pools are not an option, the correct answer is to use a step scaling policy with a lower scale-out threshold and reduce the health check grace period to speed up instance registration. But among given options, B is best because it triggers scale-out earlier and reduces the time to reach desired capacity. Reviewing: A is scheduled, not reactive; B reduces cooldown and threshold; C is predictive; D is not a thing; E changes instance type. Correct: B.

Key principle: NAT direction and interface roles matter as much as the IP address mapping. Inside/outside designation controls which traffic is translated.