After deploying a new version of your service, you notice an unexpected increase in the number of instances in your Kubernetes cluster. The service is configured to scale when the average CPU utilization exceeds 80%. Which tool would be most effective for diagnosing this issue?