Vertical Pod Autoscaler Deep Dive

💡 Quick Answer: Deploy VPA in Off mode first to get recommendations without changes, then switch to Auto for memory only (let HPA handle CPU scaling). Set minAllowed and maxAllowed to prevent extreme values. VPA and HPA can coexist if they control different resources.

The Problem

Developers guess resource requests/limits at deploy time — usually too high (wasting cluster capacity) or too low (causing OOM kills and throttling). VPA observes actual usage and adjusts requests automatically, but misconfiguration can cause pod restarts or conflict with HPA.

The Solution

Step 1: Recommendation Mode (Safe Start)

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

Check recommendations:

kubectl describe vpa my-app-vpa
# Recommendation:
#   Container: app
#     Lower Bound:  Cpu: 25m, Memory: 128Mi
#     Target:       Cpu: 100m, Memory: 256Mi
#     Upper Bound:  Cpu: 500m, Memory: 1Gi
#     Uncapped:     Cpu: 100m, Memory: 256Mi

Step 2: Auto Mode with Bounds

apiVersion: autoscaling.k8s.io/v1
kind: VerticalPodAutoscaler
metadata:
  name: my-app-vpa
spec:
  targetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: my-app
  updatePolicy:
    updateMode: "Auto"
  resourcePolicy:
    containerPolicies:
      - containerName: app
        minAllowed:
          cpu: 50m
          memory: 128Mi
        maxAllowed:
          cpu: "4"
          memory: 8Gi
        controlledResources: ["cpu", "memory"]

VPA + HPA Coexistence

# VPA controls memory only
apiVersion: autoscaling.k8s.io/v1
kind: VerticalPodAutoscaler
metadata:
  name: my-app-vpa
spec:
  targetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: my-app
  updatePolicy:
    updateMode: "Auto"
  resourcePolicy:
    containerPolicies:
      - containerName: app
        controlledResources: ["memory"]
        minAllowed:
          memory: 128Mi
        maxAllowed:
          memory: 8Gi
---
# HPA controls CPU scaling (replica count)
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: my-app-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: my-app
  minReplicas: 2
  maxReplicas: 10
  metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 70

graph TD
    VPA[VPA<br/>controlledResources: memory] -->|Adjust memory<br/>requests/limits| POD[Pod]
    HPA[HPA<br/>metric: CPU utilization] -->|Scale replicas| DEPLOY[Deployment<br/>replicas: 2-10]
    DEPLOY --> POD
    
    VPA -.->|❌ Does NOT control| CPU[CPU resources]
    HPA -.->|❌ Does NOT control| MEM[Memory resources]

Update Modes

Common Issues

VPA keeps restarting pods

VPA evicts pods to apply new resource values (in-place resize is not yet GA). Use updateMode: Initial to only set resources at pod creation.

VPA and HPA fighting over CPU

Never let both VPA and HPA control the same resource. Use controlledResources: ["memory"] on VPA when HPA targets CPU.

VPA recommends extremely low values

Set minAllowed to prevent VPA from under-provisioning. This is common for services with spiky traffic.

Best Practices

• Start with Off mode — observe recommendations for a week before enabling Auto
• Set minAllowed and maxAllowed — prevent extreme right-sizing
• VPA for memory, HPA for CPU — the safe coexistence pattern
• Use Initial mode for StatefulSets — avoid evicting database pods
• Monitor VPA events — kubectl get events --field-selector reason=EvictedByVPA

Key Takeaways

• VPA right-sizes resource requests based on observed usage
• Start in Off mode to see recommendations without changes
• VPA + HPA coexistence: VPA controls memory, HPA controls CPU/replicas
• minAllowed/maxAllowed are essential guardrails
• VPA evicts pods to apply changes (no in-place resize yet in GA)
• Initial mode is safest for stateful workloads