Scaling

Lucity supports two scaling modes: fixed replica counts and automatic horizontal pod autoscaling (HPA). Scaling is configured per service per environment, so you can run a single replica in development and autoscale in production.

Manual scaling

Set a fixed number of replicas for a service:

mutation {
  setServiceScaling(input: {
    projectId: "myapp"
    environment: "production"
    service: "api"
    replicas: 3
  }) {
    replicas
  }
}

The replica count must be between 1 and 20. Lucity updates the Helm values in the GitOps repo and ArgoCD applies the change to Kubernetes.

Autoscaling

Enable horizontal pod autoscaling to let Kubernetes adjust replicas based on CPU utilization:

mutation {
  setServiceScaling(input: {
    projectId: "myapp"
    environment: "production"
    service: "api"
    replicas: 2
    autoscaling: {
      enabled: true
      minReplicas: 2
      maxReplicas: 10
      targetCPU: 70
    }
  }) {
    replicas
    autoscaling {
      enabled
      minReplicas
      maxReplicas
      targetCPU
    }
  }
}

When autoscaling is enabled, Kubernetes monitors CPU usage across your pods. If average CPU exceeds the target percentage, new replicas are added (up to maxReplicas). When load drops, replicas scale back down (to minReplicas).

Constraints

Per-environment scaling

Scaling is configured independently per environment. A typical setup:

• Development: 1 replica, no autoscaling. Keep it simple.
• Staging: 1-2 replicas. Enough to test, not enough to burn budget.
• Production: autoscaling with 2-10 replicas at 70% target CPU. Let Kubernetes handle the load.

How it maps to Kubernetes

Manual scaling sets the replicas field on the Deployment. Autoscaling creates a HorizontalPodAutoscaler resource targeting the Deployment.

Both configurations live in the GitOps repo as Helm values. After ejection, scaling continues working as standard Kubernetes Deployment replicas and HPA resources.