Question

A media company runs a fleet of EC2 instances behind an Application Load Balancer in an Auto Scaling group. Deployments use AWS CodeDeploy with an in-place deployment strategy that runs configuration scripts on each existing instance during release. Recently, deployments have caused inconsistent runtime state: some instances end up with drifted package versions, and failed deployments leave the fleet in a partially updated state that is difficult to roll back. Operations wants to eliminate configuration drift, guarantee that every instance runs an identical, tested artifact, and enable fast, reliable rollback. Which approach BEST meets these requirements while improving operational excellence?

Accepted Answer

Build immutable golden AMIs in the CI/CD pipeline with EC2 Image Builder, then deploy new versions by launching a fresh Auto Scaling group from the new AMI using CodeDeploy blue/green with automatic rollback on alarm.

Answer

Continue in-place deployments but add a CodeDeploy lifecycle hook that runs a script to verify and force-correct package versions on each instance after each deployment.

Answer

Replace CodeDeploy with AWS Systems Manager State Manager associations that continuously reapply the desired configuration to running instances on a schedule to prevent drift.

Answer

Enable Auto Scaling instance refresh with a longer health check grace period and keep the existing in-place CodeDeploy scripts so instances are gradually replaced during deployments.