Question

A DevOps team manages a production CloudFormation stack that includes an Amazon RDS DB instance and several Lambda functions. Recent template updates have occasionally caused unexpected resource replacements, resulting in unplanned downtime. The team wants a process that lets them see exactly which resources will be added, modified, or replaced before any changes are applied, and requires an explicit approval step before execution. Which approach BEST meets these requirements?

Accepted Answer

Create a CloudFormation change set from the updated template, review the proposed actions (including the Replacement column), and execute the change set only after approval.

Answer

Enable termination protection on the stack and run update-stack directly, relying on the rollback-on-failure behavior to undo any unwanted replacements.

Answer

Run CloudFormation drift detection before each update-stack call to identify which resources will be replaced by the deployment.

Answer

Add a DeletionPolicy of Retain to all resources so that any replacement preserves the original resource instead of causing downtime.