Expose actual pod CPU/memory request from status.containerStatuses.resources (Kubernetes 1.33) #2665
Description
What would you like to be added:
Expose a new set of metrics in kube-state-metrics that report the actual CPU and memory requests and limits from status.conatinerStatuses[i].resources (introduced in Kubernetes 1.33).
Why is this needed:
Starting from Kubernetes 1.33, conatiners always report their actual active resource requests and limits under status.conatinerStatuses[i].resources.
Currently, kube-state-metrics only exposes values from spec.resources, which reflect the desired future state rather than the actual current.
Describe the solution you'd like
Add new metrics like:
- kube_pod_container_actual_resource_requests,
- kube_pod_container_actual_resource_limits
These metrics should expose values from status.conatinerStatuses[i].resources and reflect the current actual CPU and memory requests and limits of each container.
Additional context