Question

A DevOps engineer is updating an existing service in an Amazon Elastic Container Service (Amazon ECS) cluster. The service uses the rolling update deployment type. All cluster updates are deployed by using AWS CloudFormation. Currently, no automated remediation takes place when a deployment fails. The service stays in a FAILED state until the DevOps team manually resolves the identified issues. The DevOps engineer must configure an automatic solution to remediate a deployment failure caused by the service task failing to start. The service must automatically roll back to its most recently completed deployment. Which solution meets these requirements with the **LEAST** operational overhead?

Accepted Answer

C. Update the service to use the ECS deployment circuit breaker failure detection support. Configure the service to roll back on failures.

Answer

Configure Amazon CloudWatch alarms that check the service's runtime metrics. Update the service deployment configuration to reference the new alarms.

Answer

Set an Amazon EventBridge rule that uses the SERVICE_DEPLOYMENT_FAILED event name. Add an AWS Lambda function target that rolls the service back to the most recent deployment that is in a COMPLETED state.

Answer

Use an AWS CodeDeploy deployment group for all ECS service updates. Configure automatic rollbacks on the deployment group to roll back to the most recent successful deployment when a deployment fails.