Gke Scaling

GKE Workload Scaling

This reference covers scaling workloads on GKE. The golden path enables VPA, OPTIMIZE_UTILIZATION autoscaling profile, and Node Auto Provisioning by default.

> MCP Tools: get_k8s_resource, describe_k8s_resource, > apply_k8s_manifest, patch_k8s_resource, get_cluster, update_cluster, > update_node_pool

Golden Path Scaling Defaults

Setting | Golden Path Value | Notes ---------------------------------------- | ---------------------- | ----- autoscaling.autoscalingProfile | OPTIMIZE_UTILIZATION | Aggressive scale-down for cost savings verticalPodAutoscaling.enabled | true | VPA recommendations available autoscaling.enableNodeAutoprovisioning | true | NAP creates node pools on demand GPU resource limits (T4, A100) | 1000000000 each | NAP can provision GPU nodes

Scaling Mechanisms

1. Manual Scaling

> kubectl-only — no MCP equivalent for kubectl scale. Use kubectl > directly.

kubectl scale deployment  --replicas= -n 

2. Horizontal Pod Autoscaling (HPA)

Scales the number of pods based on metrics.

Quick setup (kubectl-only — no MCP equivalent for kubectl autoscale):

kubectl autoscale deployment  --cpu-percent=50 --min=1 --max=10

Manifest approach (recommended — use MCP apply_k8s_manifest):

See [assets/hpa-example.yaml](./assets/hpa-example.yaml) for a template.

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: -hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: 
  minReplicas: 1
  maxReplicas: 10
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 50

3. Vertical Pod Autoscaling (VPA)

Adjusts CPU and memory requests to match actual usage. Enabled by default on golden path.

Update modes:

• Off — recommendations only (safest, start here)
• Initial — sets resources only at pod creation
• Auto — restarts pods to apply new resource values
• InPlaceOrRecreate — updates resources without restart when possible (GKE

1.34+)

Create VPA in recommendation mode:

apiVersion: autoscaling.k8s.io/v1
kind: VerticalPodAutoscaler
metadata:
  name: -vpa
spec:
  targetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: 
  updatePolicy:
    updateMode: "Off"

Read recommendations (prefer MCP describe_k8s_resource):

# MCP (preferred)
describe_k8s_resource(parent="...", resourceType="verticalpodautoscaler", name="-vpa", namespace="")

# kubectl fallback
kubectl get vpa -vpa -o jsonpath='{.status.recommendation}'

See [assets/vpa-example.yaml](./assets/vpa-example.yaml) for a full template.

4. Cluster Autoscaler / Node Auto Provisioning (NAP)

On Autopilot (golden path), node scaling is fully managed. NAP automatically creates and sizes node pools based on workload demands.

For Standard clusters:

# Enable cluster autoscaler on a node pool
gcloud container clusters update  --region  \
  --enable-autoscaling --node-pool  \
  --min-nodes  --max-nodes  \
  --quiet

# Enable NAP
gcloud container clusters update  --region  \
  --enable-autoprovisioning \
  --min-cpu  --max-cpu  \
  --min-memory  --max-memory  \
  --quiet

Autoscaling profiles:

| Profile | Behavior | Golden Path? | | ---------------------- | ------------------------------------ | ------------ | | BALANCED | Default GKE; conservative scale-down | No | | OPTIMIZE_UTILIZATION | Aggressive scale-down; lower idle | Yes | : : resources : :

Best Practices

1. Define resource requests: HPA and VPA rely on accurate requests. Always

set them.

1. Avoid metric conflicts: Do not use HPA and VPA on the same metric.

Typical pattern: HPA on CPU, VPA on memory.

1. Pod Disruption Budgets: Define PDBs for all production workloads to

ensure availability during scaling events.

1. HPA stabilization: HPA has a default 5-minute stabilization window. Tune

behavior for faster response if needed.

1. VPA "Auto" caution: Auto mode restarts pods. Ensure your app handles

SIGTERM gracefully. VPA requires at least 2 replicas for evictions by default.

1. Use ComputeClasses: For workload-specific node targeting (Spot fallback,

GPU, specific machine families), use ComputeClasses instead of node selectors.

Rightsizing Workflow

1. Deploy VPA in Off mode for 24+ hours
2. Read recommendations: kubectl describe vpa
3. Compare target values against current requests
4. Apply with 20% buffer: new_request = target * 1.2
5. Use patch format to update Deployment

Condition | Recommendation | Risk ----------------------------- | ------------------------------------ | ------ CPU request >5x P95 actual | Reduce to P95 * 1.2 | Medium Memory request >3x P95 actual | Reduce to P95 * 1.2 | Medium CPU request >2x P95 actual | Rightsizing with 20% buffer | Low No resource limits set | Add limits to prevent noisy-neighbor | Low

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: google
• Source: google/skills
• License: Apache-2.0

Install and usage instructions live in the source repository linked above.