Deployment Helm

Helm Deployment Guide

This guide covers deploying nself-chat using Helm charts.

Table of Contents

• Prerequisites
• Quick Start
• Chart Overview
• Installation
• Configuration
• Upgrading
• Rollback
• Uninstallation
• Customization
• Troubleshooting

Prerequisites

• Kubernetes cluster (1.25+)
• Helm 3.x
• kubectl configured with cluster access
• Container registry access

Quick Start

# Deploy with defaults
./scripts/helm-deploy.sh

# Deploy to staging
./scripts/helm-deploy.sh --env staging

# Deploy to production
./scripts/helm-deploy.sh --env production --tag v1.0.0

# Preview changes
./scripts/helm-deploy.sh --dry-run

Chart Overview

deploy/helm/nself-chat/
├── Chart.yaml              # Chart metadata
├── values.yaml             # Default values
├── values-production.yaml  # Production overrides
├── values-staging.yaml     # Staging overrides
└── templates/
    ├── _helpers.tpl        # Template helpers
    ├── deployment.yaml     # Deployment template
    ├── service.yaml        # Service template
    ├── ingress.yaml        # Ingress template
    ├── configmap.yaml      # ConfigMap template
    ├── secrets.yaml        # Secrets template
    ├── hpa.yaml           # HPA template
    ├── pdb.yaml           # PDB template
    ├── serviceaccount.yaml # ServiceAccount template
    └── networkpolicy.yaml  # NetworkPolicy template

Chart Dependencies

Optional dependencies (disabled by default):

dependencies:
  - name: postgresql
    version: '13.x.x'
    repository: https://charts.bitnami.com/bitnami
    condition: postgresql.enabled
  - name: redis
    version: '18.x.x'
    repository: https://charts.bitnami.com/bitnami
    condition: redis.enabled

Installation

From Local Chart

# Update dependencies
helm dependency update ./deploy/helm/nself-chat

# Install
helm install nself-chat ./deploy/helm/nself-chat \
  --namespace nself-chat \
  --create-namespace

# Install with custom values
helm install nself-chat ./deploy/helm/nself-chat \
  --namespace nself-chat \
  --create-namespace \
  --values ./deploy/helm/nself-chat/values.yaml \
  --values ./deploy/helm/nself-chat/values-production.yaml \
  --set image.tag=v1.0.0

From OCI Registry

# Login to registry
helm registry login ghcr.io -u <username>

# Install from OCI
helm install nself-chat oci://ghcr.io/nself/charts/nself-chat \
  --version 0.2.0 \
  --namespace nself-chat \
  --create-namespace

With Environment Values

# Staging
helm install nself-chat ./deploy/helm/nself-chat \
  -f ./deploy/helm/nself-chat/values.yaml \
  -f ./deploy/helm/nself-chat/values-staging.yaml \
  --namespace nself-chat-staging \
  --create-namespace

# Production
helm install nself-chat ./deploy/helm/nself-chat \
  -f ./deploy/helm/nself-chat/values.yaml \
  -f ./deploy/helm/nself-chat/values-production.yaml \
  --namespace nself-chat \
  --create-namespace

Configuration

Key Values

Parameter	Description	Default
replicaCount	Number of replicas	2
image.repository	Image repository	ghcr.io/nself/nself-chat
image.tag	Image tag	Chart appVersion
image.pullPolicy	Pull policy	IfNotPresent
service.type	Service type	ClusterIP
service.port	Service port	80
ingress.enabled	Enable ingress	true
ingress.hosts	Ingress hosts	[{host: chat.nself.io}]
resources.requests.cpu	CPU request	100m
resources.requests.memory	Memory request	256Mi
autoscaling.enabled	Enable HPA	true
autoscaling.minReplicas	Min replicas	2
autoscaling.maxReplicas	Max replicas	10

External Services

Configure external service URLs:

externalServices:
  graphql:
    url: https://api.example.com/v1/graphql
  auth:
    url: https://auth.example.com
  storage:
    url: https://storage.example.com

Secrets

Option 1: Use existing secret (recommended):

secrets:
  create: false
  existingSecret: my-existing-secret

Option 2: Create from values (not recommended for production):

secrets:
  create: true
  hasuraAdminSecret: 'change-me'
  hasuraJwtSecret: '{"type":"HS256","key":"..."}'
  postgresPassword: 'change-me'
  redisPassword: 'change-me'

Ingress

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

Resources

resources:
  requests:
    cpu: 100m
    memory: 256Mi
  limits:
    cpu: 1000m
    memory: 1Gi

Autoscaling

autoscaling:
  enabled: true
  minReplicas: 2
  maxReplicas: 10
  targetCPUUtilizationPercentage: 70
  targetMemoryUtilizationPercentage: 80

Upgrading

Standard Upgrade

helm upgrade nself-chat ./deploy/helm/nself-chat \
  --namespace nself-chat \
  --reuse-values \
  --set image.tag=v1.1.0

Upgrade with New Values

helm upgrade nself-chat ./deploy/helm/nself-chat \
  --namespace nself-chat \
  -f ./deploy/helm/nself-chat/values.yaml \
  -f ./deploy/helm/nself-chat/values-production.yaml \
  --set image.tag=v1.1.0

Atomic Upgrade

Automatically rollback on failure:

helm upgrade nself-chat ./deploy/helm/nself-chat \
  --namespace nself-chat \
  --atomic \
  --timeout 300s

Check Upgrade Status

helm status nself-chat -n nself-chat
kubectl rollout status deployment/nself-chat -n nself-chat

Rollback

View History

helm history nself-chat -n nself-chat

Rollback to Previous

helm rollback nself-chat -n nself-chat

Rollback to Specific Revision

helm rollback nself-chat 3 -n nself-chat

Using Rollback Script

./scripts/rollback.sh --helm --namespace nself-chat --revision 3

Uninstallation

# Uninstall release
helm uninstall nself-chat -n nself-chat

# Delete namespace (optional)
kubectl delete namespace nself-chat

# Delete PVCs (if using persistent storage)
kubectl delete pvc -n nself-chat --all

Customization

Custom Values File

Create a custom values file:

# my-values.yaml
app:
  name: 'My Chat'
  environment: production

image:
  tag: v1.0.0

replicaCount: 3

ingress:
  hosts:
    - host: chat.mycompany.com
      paths:
        - path: /
          pathType: Prefix

resources:
  requests:
    cpu: 200m
    memory: 512Mi
  limits:
    cpu: 2000m
    memory: 2Gi

Install with custom values:

helm install nself-chat ./deploy/helm/nself-chat \
  -f my-values.yaml \
  --namespace nself-chat

Template Customization

To customize templates, copy and modify:

# Export current templates
helm template nself-chat ./deploy/helm/nself-chat > manifests.yaml

# Modify as needed
# Apply directly with kubectl
kubectl apply -f manifests.yaml

Adding Extra Resources

Add extra templates in templates/:

# templates/extra-configmap.yaml
{{- if .Values.extraConfig }}
apiVersion: v1
kind: ConfigMap
metadata:
  name: {{ include "nself-chat.fullname" . }}-extra
data:
  {{- toYaml .Values.extraConfig | nindent 2 }}
{{- end }}

Hooks

Add lifecycle hooks:

# templates/hooks/pre-upgrade-job.yaml
apiVersion: batch/v1
kind: Job
metadata:
  name: {{ include "nself-chat.fullname" . }}-pre-upgrade
  annotations:
    "helm.sh/hook": pre-upgrade
    "helm.sh/hook-delete-policy": hook-succeeded
spec:
  template:
    spec:
      containers:
        - name: pre-upgrade
          image: {{ include "nself-chat.image" . }}
          command: ["npm", "run", "db:migrate"]
      restartPolicy: Never

Troubleshooting

Helm Commands

# Check release status
helm status nself-chat -n nself-chat

# Get release values
helm get values nself-chat -n nself-chat

# Get release manifest
helm get manifest nself-chat -n nself-chat

# Test chart rendering
helm template nself-chat ./deploy/helm/nself-chat --debug

Common Issues

Dependency Update Failed

# Clear cache and retry
rm -rf ./deploy/helm/nself-chat/charts ./deploy/helm/nself-chat/Chart.lock
helm dependency update ./deploy/helm/nself-chat

Template Rendering Errors

# Debug template rendering
helm template nself-chat ./deploy/helm/nself-chat --debug 2>&1 | head -100

Installation Timeout

# Increase timeout
helm install nself-chat ./deploy/helm/nself-chat \
  --timeout 600s \
  --wait

Values Not Applied

# Check computed values
helm get values nself-chat -n nself-chat -a

# Re-apply with explicit values
helm upgrade nself-chat ./deploy/helm/nself-chat \
  --namespace nself-chat \
  -f values.yaml \
  --force

Debug Mode

# Enable debug output
helm install nself-chat ./deploy/helm/nself-chat \
  --debug \
  --dry-run

Pod Issues

# Check pods
kubectl get pods -n nself-chat -l app.kubernetes.io/name=nself-chat

# Describe pod
kubectl describe pod <pod-name> -n nself-chat

# Check logs
kubectl logs -n nself-chat -l app.kubernetes.io/name=nself-chat

Best Practices

1. Use values files for environment-specific configuration
2. Never commit secrets to values files
3. Use semantic versioning for chart and image tags
4. Test with --dry-run before applying
5. Use --atomic for production upgrades
6. Keep chart dependencies updated
7. Document custom values in README

CI/CD Integration

GitHub Actions Example

- name: Deploy to Kubernetes
  run: |
    helm upgrade --install nself-chat ./deploy/helm/nself-chat \
      --namespace nself-chat \
      --create-namespace \
      -f ./deploy/helm/nself-chat/values.yaml \
      -f ./deploy/helm/nself-chat/values-production.yaml \
      --set image.tag=${{ github.sha }} \
      --wait \
      --atomic \
      --timeout 300s

Related Documentation

• Docker Deployment
• Kubernetes Deployment