Pod CPU and Memory Underutilization Driving Cost Waste
infoConsistently low CPU utilization and memory usage in pods indicates over-provisioned resource requests, leading to wasted node capacity and unnecessary infrastructure costs that can be optimized through right-sizing.
Monitor pod-level CPU utilization and memory usage metrics. If utilization consistently remains below 30-40% of requested resources over extended periods, pods are likely over-provisioned. Check container/cpu/core_usage_time filtered by container_name and memory usage by pod.
Scale down pod replicas using HorizontalPodAutoscaler or reduce CPU/memory requests in pod specifications. Review and adjust resource requests based on actual usage patterns. For clusters with consistently low utilization, consider reducing node count or switching to smaller node types.