ClusterResourceQuota Controller

A Kubernetes operator for implementing cluster-wide resource quotas across multiple namespaces.

Overview

The ClusterResourceQuota controller mimics the Kubernetes ResourceQuota mechanism to work across multiple namespaces. This allows administrators to define resource constraints that apply to groups of namespaces based on label selectors.

Features

• Create quotas that apply across multiple namespaces
• Select namespaces using label selectors
• Support for all standard Kubernetes resource types (pods, services, etc.)
• Support for compute resources (CPU, memory)
• Support for storage resources (PVCs)
• Automatic aggregation of resource usage across namespaces

Usage

Installation

Note: The Helm chart is the single source of truth for all manifests (CRDs, RBAC, webhooks, cert-manager, etc.). The config/ folder and Kubebuilder-generated manifests are not used for deployment or testing. Kubebuilder markers in the Go code are for documentation only.

Install the controller using Helm:

helm install pac-quota-controller oci://ghcr.io/powerhome/pac-quota-controller-chart --version <version> -n pac-quota-controller-system --create-namespace

To upgrade:

helm upgrade pac-quota-controller oci://ghcr.io/powerhome/pac-quota-controller-chart --version <version> -n pac-quota-controller-system

End-to-End (e2e) Testing

All e2e tests use Helm for deployment. The config/ folder is ignored and not used for testing or production. To run e2e tests:

# Install chart for testing
helm install pac-quota-controller ./charts/pac-quota-controller -n pac-quota-controller-system --create-namespace
# Run tests
make test-e2e
# Uninstall after tests
helm uninstall pac-quota-controller -n pac-quota-controller-system

Example Usage

Create a ClusterResourceQuota to limit resources across namespaces:

apiVersion: quota.powerapp.cloud/v1alpha1
kind: ClusterResourceQuota
metadata:
  name: team-quota
spec:
  namespaceSelector:
    matchLabels:
      team: frontend
  hard:
    pods: "50"
    requests.cpu: "10"
    requests.memory: 20Gi
    limits.cpu: "20"
    limits.memory: 40Gi
    # Storage resources
    requests.storage: "100Gi"                    # PVC storage requests
    requests.ephemeral-storage: "20Gi"           # Pod ephemeral storage requests

This quota will apply to all namespaces with the label team: frontend and limit:

• Total pods to 50
• Total CPU requests to 10 cores and limits to 20 cores
• Total memory requests to 20Gi and limits to 40Gi
• Total storage requests from PVCs to 100Gi
• Total ephemeral storage requests from Pods to 20Gi

Contributing

See CONTRIBUTING.md.