Validate HPA Autoscaling Configuration in Helm

Horizontal Pod Autoscaler (HPA) Configuration

Helm charts commonly include an autoscaling section that controls HPA behavior. Misconfigured autoscaling can cause over-provisioning (wasting money) or under-provisioning (causing outages).

Standard Pattern

autoscaling:
  enabled: true
  minReplicas: 2
  maxReplicas: 10
  targetCPUUtilizationPercentage: 80
  # targetMemoryUtilizationPercentage: 80  # optional

What Gets Validated

1. minReplicas <= maxReplicas: Validates that minimum is not greater than maximum.
2. Type correctness: enabled must be boolean, replica counts must be numbers.
3. replicaCount conflict: When autoscaling is enabled and replicaCount is also set, the validator notes that HPA will override the static replica count.
4. Resource requirements: HPA requires resources.requests to be set for CPU/memory-based scaling.

Common Mistakes

min > max (error):

autoscaling:
  enabled: true
  minReplicas: 10
  maxReplicas: 5  # ERROR: min > max

Autoscaling enabled with static replicaCount:

replicaCount: 3  # This will be ignored when HPA is active
autoscaling:
  enabled: true
  minReplicas: 2
  maxReplicas: 10

Missing resource requests (HPA won't work):

autoscaling:
  enabled: true
  targetCPUUtilizationPercentage: 80
resources: {}  # No CPU requests = HPA can't calculate utilization

Best Practices

• Set minReplicas to at least 2 for high availability
• Set targetCPUUtilizationPercentage between 60-80%
• Always define resources.requests.cpu when using CPU-based HPA
• Remove or comment out replicaCount when autoscaling is enabled