DOP-C02 Practice Question 70: FIS Experiment Stop Condition
AWS Certified DevOps Engineer – Professional (DOP-C02) · Test 7 · question 10 of 10
A DevOps team is designing a chaos engineering experiment with AWS Fault Injection Service to terminate a percentage of EC2 instances in an Auto Scaling group during business hours. Leadership requires that if the experiment pushes the service's error rate above an agreed threshold, the experiment halts automatically before customers are meaningfully affected, without anyone needing to watch a dashboard and cancel it manually. How should the team configure the FIS experiment template to satisfy this requirement?
Show answer and explanation
Answer:
An FIS experiment template supports one or more stop conditions, each pointing at a CloudWatch alarm; if the referenced alarm enters the ALARM state while the experiment is running, FIS automatically stops the experiment without any human action. This is the built-in, automated safety mechanism the requirement describes. Option A limits how long the disruptive action runs, but a fixed duration is blind to the actual error rate and could still cause meaningful customer impact within that window, or stop too early to be useful. Option C limits blast radius by choosing fewer targets, which reduces risk but still does not react to real-time error rate signals and does not guarantee an automatic halt if the error rate spikes anyway. Option D relies on a human noticing the email and manually stopping the experiment, which is exactly the manual dependency leadership wants removed.
Reference: Stop conditions for AWS FIS