For Chinese version, see 中文

This is a Helm Chart for deploying Dify on Azure Kubernetes Service (AKS). Dify is an open-source LLM application development platform, and this Helm Chart aims to simplify its deployment and management in Kubernetes environments.

• 🚀 One-click deployment of complete Dify platform to AKS
• 🔒 Built-in security configuration and secret management
• 🔄 CI/CD automation deployment support
• 📦 Includes all necessary dependencies (Redis, PostgreSQL, MinIO, etc.)
• 🔍 Multiple vector database options (Weaviate, Qdrant)

This Helm Chart deploys the Dify platform with the following core components:

• Frontend: Dify's web interface
• API Server: Backend API service
• Worker: Asynchronous task processing service
• PostgreSQL: Main database (v12.1.6)
• Redis: Cache and message queue (v18.1.2)
• MinIO: Object storage service (v12.8.7)
• Vector Database: Vector database (supports Weaviate v16.3.0 or Qdrant v0.7.0)

Architecture diagram sourced from Dify official Docker Compose deployment

• Azure Kubernetes Service (AKS) cluster
• Azure DevOps account (for CI/CD)
• Helm 3.x
• kubectl configuration

1. Create a custom values file:

global:
  # Host configuration
  # Supports two configuration methods:
  # 1. IP address: For internal network access, e.g., "10.104.179.207"
  # 2. Domain name: For public network access, e.g., "dify.your-domain.com"
  # Note: When using domain names, ensure DNS resolution is correct and configure TLS certificates
  host: "10.104.179.207"  # Use frontend LoadBalancer IP as hostname
  
  # Port configuration
  # Usually no need to set, use default port
  # If custom port is needed, ensure it matches LoadBalancer configuration
  port: ""  # No need to set port
  
  # TLS configuration
  # - false: For internal network access or HTTP access
  # - true: For public network access, requires TLS certificate configuration
  enableTLS: false  # Internal network access doesn't need TLS
  
  image:
    tag: "1.4.1"  # Currently supported Dify version
  
  edition: "SELF_HOSTED"
  storageType: "s3"  # Use S3 storage

  # Additional environment variables configuration
  extraBackendEnvs:
  - name: SECRET_KEY
    valueFrom:
      secretKeyRef:
        name: dify-secrets
        key: SECRET_KEY
  - name: VECTOR_STORE
    value: "weaviate"  # or "qdrant"

# Component configuration
postgresql:
  enabled: true
  # PostgreSQL configuration...

redis:
  enabled: true
  # Redis configuration...

minio:
  enabled: true
  # MinIO configuration...

weaviate:
  enabled: true
  # Weaviate configuration...

qdrant:
  enabled: false
  # Qdrant configuration...

2. Push Helm Chart to Azure DevOps repository and execute Pipeline:

# Push Helm Chart to Azure DevOps repository
git add .
git commit -m "feat: update Dify Helm Chart configuration"
git push origin main

# Pipeline will automatically execute the following steps:
# 1. Create Kubernetes namespace
# 2. Generate and create necessary secrets
# 3. Update Helm dependencies
# 4. Deploy Dify application
# 5. Verify deployment status
# 6. Wait for services to be ready

After Pipeline execution completes, you will see output similar to:

================================================
Dify deployment completed!
Frontend access address: http://10.104.179.207
================================================

• Chart version: 0.1.0
• Dify application version: 1.4.1
• Dependency versions:
  • Redis: 18.1.2
  • PostgreSQL: 12.1.6
  • MinIO: 12.8.7
  • Weaviate: 16.3.0
  • Qdrant: 0.7.0

This project includes Azure DevOps Pipeline configuration (azure-pipeline-helm.yaml) that supports automated deployment. Main features include:

• Automatic namespace creation
• Secret generation and management
• Helm dependency updates
• Application deployment
• Deployment status verification
• Service health checks

Configure the following variables in Azure DevOps:

azureSubscriptionEndpoint: 'your-azure-subscription'
azureResourceGroup: 'your-resource-group'
kubernetesCluster: 'your-aks-cluster'
namespace: 'dify'
helmReleaseName: 'dify'

• Use Kubernetes Secrets to manage sensitive information
• Configure appropriate resource limits and requests
• Enable network policies
• Regularly update keys and certificates

Strongly recommend using external secure storage services:

• Database: Use Azure Database for PostgreSQL or Azure Database for MySQL

  • Provides automatic backup and recovery
  • Built-in security and compliance
  • Auto-scaling and high availability
  • Managed maintenance and updates

• Object Storage: Use Azure Blob Storage or Azure Files

  • Provides data redundancy and disaster recovery
  • Built-in encryption and access control
  • Auto-scaling storage capacity
  • Better cost efficiency

• Cache: Use Azure Cache for Redis

  • Provides high-performance caching service
  • Built-in security and monitoring
  • Auto-scaling and failover

Benefits of using external storage:

1. Data Security: Professional data protection and backup strategies
2. High Availability: 99.9%+ service availability guarantee
3. Auto-scaling: Automatically adjust resources based on demand
4. Cost Optimization: Pay-as-you-use, avoid resource waste
5. Simplified Operations: Reduce operational burden, focus on application development

Configuration example:

# Disable built-in database, use external PostgreSQL
postgresql:
  enabled: false

# Disable built-in Redis, use external Redis
redis:
  enabled: false

# Disable built-in MinIO, use external storage
minio:
  enabled: false

# Configure external storage connection information
global:
  extraBackendEnvs:
  - name: DB_HOST
    value: "your-postgresql-server.postgres.database.azure.com"
  - name: DB_PORT
    value: "5432"
  - name: DB_DATABASE
    value: "dify"
  - name: DB_USERNAME
    value: "dify_user"
  - name: DB_PASSWORD
    valueFrom:
      secretKeyRef:
        name: dify-secrets
        key: DB_PASSWORD
  - name: REDIS_HOST
    value: "your-redis-server.redis.cache.windows.net"
  - name: REDIS_PORT
    value: "6380"
  - name: REDIS_PASSWORD
    valueFrom:
      secretKeyRef:
        name: dify-secrets
        key: REDIS_PASSWORD
  - name: S3_ENDPOINT
    value: "https://your-storage-account.blob.core.windows.net"
  - name: S3_BUCKET_NAME
    value: "dify-storage"
  - name: S3_ACCESS_KEY
    valueFrom:
      secretKeyRef:
        name: dify-secrets
        key: S3_ACCESS_KEY
  - name: S3_SECRET_KEY
    valueFrom:
      secretKeyRef:
        name: dify-secrets
        key: S3_SECRET_KEY

• Configure multiple replicas
• Use Pod anti-affinity
• Configure appropriate resource limits
• Use persistent storage

• Configure appropriate log levels
• Set up resource usage alerts
• Monitor service health status
• Regularly check system logs

1. Update image version in values.yaml
2. Execute Helm upgrade
3. Run database migration

helm upgrade dify dify-helm/dify -f values.yaml
kubectl exec -it $(kubectl get pod -l app.kubernetes.io/component=api -o jsonpath='{.items[0].metadata.name}') \
  -n dify -- flask db upgrade

• Regularly backup PostgreSQL database
• Backup MinIO storage data
• Save Helm values configuration

Common issues and solutions:

1. Pod startup failure

   • Check resource limits
   • Verify environment variable configuration
   • View Pod logs

2. Database connection issues

   • Verify database credentials
   • Check network policies
   • Confirm database service status

3. Storage access issues

   • Check storage class configuration
   • Verify access credentials
   • Confirm storage service status

Welcome to submit Pull Requests or create Issues. Before submitting code, please ensure:

1. Follow Helm Chart best practices
2. Update version numbers
3. Add necessary documentation
4. Pass all tests

This project is licensed under Apache License 2.0. See LICENSE file for details.

• Dify Official Documentation
• Helm Documentation
• Azure Kubernetes Service Documentation
• Azure DevOps Documentation