I’m facing a puzzling situation with resource usage in my Kubernetes namespace.

Namespace Quota:
Resource    Used     Hard
--------    ----     ----
count/pods  94       300
cpu         17800m   32
memory      66949Mi  192Gi

The namespace is reporting 17800m (17.8 cores) of CPU request quota usage. However, when I sum the CPU requests of all the pods in the namespace, the total is significantly lower (around 12 cores). I don’t understand the source of this discrepancy.

Things I’ve checked:

Horizontal Pod Autoscalers (HPA): No HPAs are active in the namespace.

Pod Requests: All pods have CPU requests defined.

What could be causing the namespace CPU quota usage to exceed the sum of the pod CPU requests?
Are there any other factors or hidden components that could be consuming CPU resources?