A DevOps Engineer must track the health of a stateless RESTful service sitting behind a Classic Load Balancer. The deployment of new application revisions is through a CI/CD pipeline. If the service's latency increases beyond a defined threshold, deployment should be stopped until the service has recovered.
Which of the following methods allow for the QUICKEST detection time?

A. Use Amazon CloudWatch metrics provided by Elastic Load Balancing to calculate average latency. Alarm and stop deployment when latency increases beyond the defined threshold. B. Use AWS Lambda and Elastic Load Balancing access logs to detect average latency. Alarm and stop deployment when latency increases beyond the defined threshold. C. Use AWS CodeDeploy's MinimumHealthyHosts setting to define thresholds for rolling back deployments. If these thresholds are breached, roll back the deployment. D. Use Metric Filters to parse application logs in Amazon CloudWatch Logs. Create a filter for latency. Alarm and stop deployment when latency increases beyond the defined threshold.