This is a dedicated watch page for a single video.
A company has an Amazon ECS cluster with Service Auto Scaling which consists of multiple Amazon EC2 instances that runs a Docker-based application. The development team always pushes a new image to a private Docker container registry whenever they publish a new version of their application. They stop and start all of the tasks to ensure that the containers have the latest version of the application. However, the team noticed that the new tasks are occasionally running the old image of the application. As the DevOps engineer, what should you do to fix this issue?