You manage a three-tier web application consisting of an autoscaled web proxy tier, an autoscaled application tier, and an Amazon RDS database tier. You use a load balancer to distribute requests from end users to the web proxy tier and another, internal load balancer to distribute requests between the web tier and the application tier. After deploying a small database schema update, you notice that all of your web and application instances have been terminated.
What may have caused this?

A. Your load balancers use an HTTP health check, and the page relies on retrieving data from your database.
B. Your load balancer use TCP health checks to provide application-level health checks.
C. The cooldown period of the Auto Scaling group is too short, so the instances do not have enough time to recover from an issue.
D. Your Auto Scaling group health check type is set to "EC2" to check that the instances themselves are healthy.