Question

A team uses AWS CodePipeline to deploy a serverless API. After the build stage produces artifacts and deploys to a dedicated 'test' environment, they want to run automated integration tests that call the deployed API endpoints. If any test fails, the pipeline must stop and NOT promote the build to the production stage. What is the most appropriate way to implement this within the pipeline?

Accepted Answer

Add a CodeBuild action in a test stage that runs the integration test suite against the deployed test environment; a non-zero exit code fails the action and halts the pipeline before production.

Answer

Add a manual approval action after deployment so a reviewer can run the tests locally and decide whether to approve promotion to production.

Answer

Configure the CodeDeploy deployment group in the test stage to run the integration tests as an AfterAllowTraffic lifecycle hook and rely on CloudWatch alarms to stop the pipeline.

Answer

Enable CodePipeline rollback on the production stage so that if integration tests fail after production deployment, the pipeline automatically reverts to the previous version.