Usage

Overview

KubeCloudScaler allows you to automatically scale your Kubernetes and cloud resources based on time-based rules called periods. This guide explains how to configure and use the three custom resources provided by KubeCloudScaler.

Custom Resources

All KubeCloudScaler CRDs are cluster-scoped (no namespace required) and use the kubecloudscaler.cloud/v1alpha3 API version.

CRD	Purpose	Use case
K8s	Scale Kubernetes workloads	Scale Deployments, StatefulSets, CronJobs, HPAs, GitHub ARS based on time
Gcp	Scale GCP resources	Start/stop Compute Engine VM instances based on time
Flow	Orchestrate multi-resource scaling	Coordinate scaling across multiple K8s and Gcp resources with timing delays

Common Structure

All scalers share a common structure with periods defining when to scale and resources defining what to scale:

apiVersion: kubecloudscaler.cloud/v1alpha3
kind: K8s  # or Gcp, Flow
metadata:
  name: my-scaler
spec:
  periods:
    - type: "down"          # "down" or "up"
      name: "night-scale"   # optional period name
      time:
        recurring:           # or fixed
          days: [all]
          startTime: "20:00"
          endTime: "07:00"
          timezone: "Europe/Paris"
      minReplicas: 0
  resources:
    types:
      - deployments
  config:
    # Resource-specific configuration

Key Concepts

Periods: Define when resources should be scaled (recurring or fixed time windows)
Resources: Define what types of resources to target (by type, name, or label selector)
Config: Resource-specific configuration (namespaces, authentication, etc.)
Dry-run mode: Preview scaling actions without executing them (dryRun: true)
Restore on delete: Automatically restore resources to their original state when the scaler CR is deleted