Skip to main content

Using Helm Templates

This guide covers how to work with Hiroba's Helm chart templates in detail.

Template Structure​

Every Hiroba Helm chart follows this structure:

app-template/
├── Chart.yaml          # Chart metadata
├── values.yaml         # Default configuration values
└── templates/
    ├── _helpers.tpl    # Reusable template helpers
    ├── deployment.yaml # Application deployment
    ├── service.yaml    # Service exposure
    ├── ingress.yaml    # Ingress (optional)
    └── serviceaccount.yaml

Customizing Values​

The values.yaml file is your primary configuration surface. Key sections:

Image Configuration​

image:
  repository: ghcr.io/your-org/your-app
  pullPolicy: IfNotPresent
  tag: "1.2.3"  # Pinned version — never use "latest"

Resource Limits​

Keep these reasonable for homelab hardware. These defaults work well on a Raspberry Pi or mini PC:

resources:
  limits:
    cpu: 500m
    memory: 256Mi
  requests:
    cpu: 50m
    memory: 64Mi

Enabling Ingress​

ingress:
  enabled: true
  className: nginx
  annotations:
    cert-manager.io/cluster-issuer: letsencrypt-prod
  hosts:
    - host: app.home.lab
      paths:
        - path: /
          pathType: Prefix
  tls:
    - secretName: app-tls
      hosts:
        - app.home.lab

Extending the Template​

To add custom resources, create new files in the templates/ directory. Use the helpers from _helpers.tpl for consistent naming and labeling:

apiVersion: v1
kind: ConfigMap
metadata:
  name: {{ include "app-template.fullname" . }}-config
  labels:
    {{- include "app-template.labels" . | nindent 4 }}
data:
  app.conf: |
    # your config here

Linting and Testing​

# Lint the chart
helm lint ./my-app

# Render templates without installing
helm template my-release ./my-app

# Dry-run install
helm install my-release ./my-app --dry-run