In the cloud, you architect systems that are designed to handle failure. In an event when a failure occurs the application owner should be aware of the failures and should have a specific action to respond to failure.

Ideally, you should also have an RCA (Root Cause Analysis) to prevent such failure from happening

Foundations: IAM, VPC

Change Management: AWS Cloud Trail

Failure Management: Cloud Formation, RDS multi-AZ. 

Cloud Formation, RDS multi-AZ.