A developer migrated an application to Amazon EKS and moved images to Amazon ECR. They created a new AWS account and updated the application configuration to point to new backend resources. The pipeline build and deployment succeeded, but the running application is still connecting to the old backend. Investigation shows the application configuration still references the original EKS cluster. Which reason explains why the application did not connect to the new resources?

The application is still connecting to the old backend because the EKS deployment is likely pulling an older version of the Docker image that contains the outdated configuration. When the developer updated the application configuration, they should have also built a new Docker image and assigned it a new, unique tag. The EKS deployment manifest then needs to be updated to reference this new image tag. Without updating the image tag in the deployment, EKS will continue to use the previously deployed image, which still has the old backend references. Incorrect options: "The developer did not actually create the new AWS account successfully" is unlikely, as the pipeline build and deployment succeeded, implying account access. "The developer added a new tag to the Docker image" would typically be a correct step towards resolving the issue, not causing it. "The developer pushed the changes to a new Docker image tag" would also be a correct step, meaning the application should have connected to the new resources if the EKS deployment was updated.