Guide: Dify in Production — LLM App Platform

Guide: n8n Workflow Automation in Production

n8n in queue mode with Redis, PostgreSQL persistence, and proper process supervision — the boring, reliable setup that survives restarts and workload spikes. The automation layer behind several of my

Guide: Production k3s Cluster with Rancher, Longhorn and MetalLB

A complete Kubernetes lab-to-production pattern: k3s in Docker with the embedded etcd datastore, Rancher for management, MetalLB for Layer-2 LoadBalancer IPs, Longhorn replicated storage, and Traefik

Dify — visual LLM app builder with RAG, agents, and prompt workflows — backed by PostgreSQL, Redis, Weaviate, an SSRF proxy and a code sandbox. The deployment pattern I use for LLM applications that non-developers build on.

What this is

A deployment guide with working examples — not a downloadable product. Everything below is the actual configuration I use, abridged to the parts that matter.

Deployment example

services:
  api:
    image: langgenius/dify-api:${DIFY_VERSION:-latest}
    container_name: dify-api
    restart: unless-stopped
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_started
      weaviate:
        condition: service_healthy
    volumes:
      - dify_storage_data:/app/api/storage
    ports:
      - ${DIFY_API_PORT:-5001}:5001
    environment:
      - MODE=api
      - SECRET_KEY=${DIFY_SECRET_KEY:?DIFY_SECRET_KEY is required}
      - INIT_PASSWORD=${DIFY_INIT_PASSWORD:-}
      - LOG_LEVEL=${LOG_LEVEL:-INFO}
      - DB_USERNAME=${POSTGRES_USER:-dify}
      - DB_PASSWORD=${POSTGRES_PASSWORD:?POSTGRES_PASSWORD is required}
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_DATABASE=${POSTGRES_DB:-dify}
      - REDIS_HOST=redis
      - REDIS_PORT=6379
      - REDIS_PASSWORD=${REDIS_PASSWORD:?REDIS_PASSWORD is required}
      - REDIS_DB=0
      - WEAVIATE_HOST=weaviate
      - WEAVIATE_PORT=8080
      - WEAVIATE_API_KEY=${WEAVIATE_API_KEY:?WEAVIATE_API_KEY is required}
      - WEAVIATE_BATCH_SIZE=100
      - STORAGE_TYPE=${STORAGE_TYPE:-local}
      - STORAGE_LOCAL_PATH=storage
      - CELERY_BROKER_URL=redis://:${REDIS_PASSWORD:?REDIS_PASSWORD is required}@redis:6379/1
      - SENTRY_DSN=
      - CONSOLE_API_URL=${CONSOLE_API_URL:-http://localhost:5001}
      - APP_API_URL=${APP_API_URL:-http://localhost:5001}
      - SSRF_PROXY_URL=http://ssrf_proxy:3128
    healthcheck:
      test:
        - CMD
        - curl
        - -f
        - http://localhost:5001/health
      interval: 30s
      timeout: 10s
      retries: 3
    networks:
      - dify_network
  worker:
    image: langgenius/dify-api:${DIFY_VERSION:-latest}
    container_name: dify-worker
    restart: unless-stopped
    depends_on:
      api:
        condition: service_healthy
    volumes:
      - dify_storage_data:/app/api/storage
    environment:
      - MODE=worker
      - SECRET_KEY=${DIFY_SECRET_KEY:?DIFY_SECRET_KEY is required}
      - LOG_LEVEL=${LOG_LEVEL:-INFO}
      - DB_USERNAME=${POSTGRES_USER:-dify}
      - DB_PASSWORD=${POSTGRES_PASSWORD:?POSTGRES_PASSWORD is required}
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_DATABASE=${POSTGRES_DB:-dify}
      - REDIS_HOST=redis
      - REDIS_PORT=6379
      - REDIS_PASSWORD=${REDIS_PASSWORD:?REDIS_PASSWORD is required}
      - REDIS_DB=0
      - WEAVIATE_HOST=weaviate
      - WEAVIATE_PORT=8080
      - WEAVIATE_API_KEY=${WEAVIATE_API_KEY:?WEAVIATE_API_KEY is required}
      - CELERY_BROKER_URL=redis://:${REDIS_PASSWORD:?REDIS_PASSWORD is required}@redis:6379/1
      - CELERY_WORKER_CONCURRENCY=${CELERY_WORKER_CONCURRENCY:-4}
      - CELERY_WORKER_MAX_TASKS_PER_CHILD=${CELERY_WORKER_MAX_TASKS_PER_CHILD:-100}
      - SENTRY_DSN=
      - CONSOLE_API_URL=${CONSOLE_API_URL:-http://localhost:5001}
      - APP_API_URL=${APP_API_URL:-http://localhost:5001}
      - SSRF_PROXY_URL=http://ssrf_proxy:3128
    healthcheck:
      test:
        - CMD-SHELL
        - celery -p app.celery.app inspect ping -d celery@$$HOSTNAME || exit 1
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 30s
    networks:
      - dify_network

Excerpt — api, worker from the compose file. The remaining services (proxies, init jobs, exporters) follow the same pattern and mount their configuration from a config/ directory.

Essential environment variables

Copy .env.example to .env and at minimum set:

POSTGRES_PASSWORD=<generate-strong-password>
REDIS_PASSWORD=<generate-strong-password>
WEAVIATE_API_KEY=<generate-api-key>
DIFY_SECRET_KEY=<generate-64-char-string>
DIFY_INIT_PASSWORD=
POSTGRES_USER=dify

Production checklist

• Set strong POSTGRES_PASSWORD, REDIS_PASSWORD, WEAVIATE_API_KEY, DIFY_SECRET_KEY
• Configure CONSOLE_API_URL and APP_API_URL with your domain
• Provide SSL certificates for Nginx (Let's Encrypt or custom)
• Set DIFY_INIT_PASSWORD for first-time admin setup
• Set up regular PostgreSQL backups (database or use Database Foundation Stack)
• Pin container versions to specific tags (e.g., DIFY_VERSION=1.2.0)
• Configure S3-compatible storage (STORAGE_TYPE=s3) for multi-server setups
• Set up monitoring: health check at http://localhost:5001/health

A guide, not a product. This page is deployment documentation with working examples — there is no zip, no download, no support contract. You adapt the patterns to your own environment, and you own the result.

Want this running production-grade in your infrastructure instead? I run this as part of the Self-Hosted AI in Production workshop — details on the consulting page: €1,000/day, remote or on-site, invoice as usual.