DOP-C02 Practice Question 28: EC2 Alarm Recover Action
AWS Certified DevOps Engineer – Professional (DOP-C02) · Test 3 · question 8 of 10
A legacy licensed application runs on a single Amazon EBS-backed EC2 instance. The vendor license is bound to the instance ID, and other systems reach the application by its private IP address, so neither may change. Roughly once a quarter the instance becomes unreachable because of a problem on the underlying host; the EC2 console shows the system status check failing while the instance status check still passes, and an engineer stops and starts the instance to fix it. The team wants this to happen automatically with the least operational overhead.
What should the team implement?
Show answer and explanation
Answer:
The recover action exists for this failure mode: it "automatically recovers the instance if it becomes impaired due to an underlying hardware failure or a problem that requires AWS involvement to repair," and "a recovered instance is identical to the original instance, including the instance ID, private IP addresses, Elastic IP addresses, and all instance metadata", which is what the license and the dependent systems require. The docs also state the recover action "can be used only with StatusCheckFailed_System, not with StatusCheckFailed_Instance", matching the observed symptom, and recommend two evaluation periods of one minute for recover alarms. It is a single alarm with no code to own.
Option A matches the wrong signal with the wrong action. A reboot restarts the guest operating system on the *same physical host*, which is the right response to an instance status check failure and no help at all when the host itself is impaired.
Option B changes the instance identity. Auto Scaling replaces a failed instance with a brand-new one, which gets a new instance ID and a new private IP address, breaking the license binding and every dependent system the stem describes.
Option C is slower and heavier for no benefit. AWS Health events are not emitted promptly for every host impairment, and the team would own an EventBridge rule, a Lambda function, and its IAM role to reproduce behavior a built-in alarm action already provides.
Reference: Stop, terminate, reboot, or recover an EC2 instance