Question

How do you resolve auto-scaling issues in Kubernetes or AWS?

Sometimes auto-scaling may not be predictable, particularly with fluctuating workloads. Outline how you would diagnose and rectify the issue based on an analysis of the metrics and changing the scaling policies with the help of Kubernetes Horizontal Pod Autoscaler or AWS Auto Scaling Groups. Focus on tools like Prometheus and Grafana in monitoring and fine-tuning thresholds to prevent performance degradation.