Cloudeka Resource Quota

Your allocated limits (spec) side by side with your live consumption (status). This is the object to check before deploying anything sizable, and the first place to look when a request is rejected for quota reasons.

kubectl get rquota -n <your-namespace>
kubectl get rquota -n <your-namespace> -o yaml

spec defines your allocated resource quota.

Field	Meaning
quotaBalance	Allocated prepaid balance
cpu	CPU quota, e.g. "4"
ram	RAM quota, e.g. "16Gi"
storage.<class>	Per-storage-class quota
gpu.<model>	Per-GPU-model count quota
floatingIP	Floating IP count quota
nvaie	NVAIE count quota

status.usages shows your current resource usage.

Same shape as spec (cpu/ram/storage/gpu/floatingIP/nvaie), recomputed live as your Pods, PVCs, and Floating IPs change. Compare status.usages.X against spec.X to see your remaining headroom.

Example

$ kubectl get rquota -n loft-acme-v-inference-1 -o yaml
spec:
  quotaBalance: 100000
  cpu: "4"
  ram: "16Gi"
  floatingIP: 1
  nvaie: 1
  gpu:
    NVIDIA-H100-80GB-HBM3: 1
  storage:
    storage-nvme-c1: "100Gi"
status:
  usages:
    cpu: "1500m"
    ram: "4Gi"

This is view-only. To raise a limit, contact the platform team — they push the change from their side, which then syncs into billing. An unsynced local edit will not take effect as a usable limit and may be overwritten.