An application is deployed on Amazon EC2 instances in an Auto Scaling group. The application includes a process that sometimes fails causing the application to return an error. Restarting the process resolves the issue. The DevOps team are investigating the issue. While the investigation is ongoing an engineer needs a method of identifying the issue and quickly remediating it without affecting the capacity of the Auto Scaling group. Which approach meets this requirement in the fastest possible time?