Question

A team deploys a containerized order-processing service to Amazon ECS using AWS CodeDeploy with a blue/green deployment. After production traffic shifts to the new (green) task set, a downstream error rate spikes. The team wants CodeDeploy to automatically stop the deployment and revert traffic to the original (blue) task set without manual intervention. Which configuration achieves this?

Accepted Answer

Associate a CloudWatch alarm with the CodeDeploy deployment group and enable automatic rollback on deployment failure or alarm threshold breach.

Answer

Set the ECS service deployment controller to ECS (rolling update) and configure a minimum healthy percent of 100.

Answer

Increase the CodeDeploy termination wait time so the blue task set stays available longer after the traffic shift.

Answer

Enable Amazon ECS Exec on the green tasks so operators can connect and manually restart the containers when errors occur.