backstory/docker-compose.yml

services:
  backstory:
    build:
      context: .
      dockerfile: Dockerfile
      target: backstory
    #image: backstory
    container_name: backstory
    restart: "always"
    env_file:
      - .env
    environment:
      - PRODUCTION=0
      - FRONTEND_URL=https://backstory-beta.ketrenos.com
      - MODEL_NAME=${MODEL_NAME:-qwen2.5:7b}
      - REDIS_URL=redis://redis:6379
      - REDIS_DB=0
      - SSL_ENABLED=true
    devices:
      - /dev/dri:/dev/dri
    depends_on:
      - ollama
      - redis
    networks:
      - internal
    ports:
      - 7860:7860 # gradio port for testing
      - 8912:8911 # FastAPI React server
    volumes:
      - ./cache:/root/.cache                     # Persist all models and GPU kernel cache
      - ./sessions:/opt/backstory/sessions:rw    # Persist sessions
      - ./dev-keys:/opt/backstory/keys:ro        # Developer keys
      - ./users:/opt/backstory/users:rw          # Live mount of user data
      - ./src:/opt/backstory/src:rw              # Live mount server src 
      - ./frontend/src/types:/opt/backstory/frontend/src/types # Live mount of types for pydantic->ts
      - ./venv:/opt/backstory/venv:rw            # Live mount for python venv
    cap_add: # used for running ze-monitor within container
      - CAP_DAC_READ_SEARCH # Bypass all filesystem read access checks
      - CAP_PERFMON         # Access to perf_events (vs. overloaded CAP_SYS_ADMIN)
      - CAP_SYS_PTRACE      # PTRACE_MODE_READ_REALCREDS ptrace access mode check

  backstory-prod:
    build:
      context: .
      dockerfile: Dockerfile
      target: backstory-prod
    #image: backstory
    container_name: backstory-prod
    restart: "always"
    env_file:
      - .env
    environment:
      - PRODUCTION=1
      - FRONTEND_URL=https://backstory.ketrenos.com
      - MODEL_NAME=${MODEL_NAME:-qwen2.5:7b}
      - REDIS_URL=redis://redis:6379
      - REDIS_DB=1
      - SSL_ENABLED=false
    devices:
      - /dev/dri:/dev/dri
    depends_on:
      - ollama
      - redis
    networks:
      - internal
    ports:
      - 8911:8911 # FastAPI React server
    volumes:
      - ./cache:/root/.cache                # Persist all models and GPU kernel cache
      - ./chromadb-prod:/opt/backstory/chromadb:rw    # Persist ChromaDB
      - ./sessions-prod:/opt/backstory/sessions:rw    # Persist sessions
      - ./docs-prod:/opt/backstory/docs:rw            # Live mount of RAG content
      - ./users-prod:/opt/backstory/users:rw          # Live mount of user data
      - ./frontend/deployed:/opt/backstory/frontend/deployed:ro   # Live mount built frontend
    cap_add: # used for running ze-monitor within container
      - CAP_DAC_READ_SEARCH # Bypass all filesystem read access checks
      - CAP_PERFMON         # Access to perf_events (vs. overloaded CAP_SYS_ADMIN)
      - CAP_SYS_PTRACE      # PTRACE_MODE_READ_REALCREDS ptrace access mode check

  redis:
    image: redis:7-alpine
    container_name: redis
    ports:
      - "6379:6379"
    volumes:
      - redis_data:/data
      - ./redis.conf:/usr/local/etc/redis/redis.conf
    command: redis-server /usr/local/etc/redis/redis.conf
    networks:
      - internal
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 30s
      timeout: 10s
      retries: 3

  # Optional: Redis Commander for GUI management
  redis-commander:
    image: rediscommander/redis-commander:latest
    container_name: redis-commander
    ports:
      - "8081:8081"
    environment:
      - REDIS_HOSTS=redis:redis:6379
    networks:
      - internal
    depends_on:
      - redis
    profiles:
      - tools  # Only start with --profile tools

  frontend:
    build:
      context: .
      dockerfile: Dockerfile
      target: frontend
    container_name: frontend
    #image: frontend
    restart: "always"
    env_file:
      - .env
    ports:
      - 3000:3000 # REACT expo while developing frontend
    volumes:
      - ./frontend:/opt/backstory/frontend:rw    # Live mount frontend src 
    networks:
      - internal

  # This doesn't work...
  # ollama-intel:
  #   image: intelanalytics/ipex-llm-inference-cpp-xpu:latest
  #   container_name: ollama-intel
  #   restart: unless-stopped
  #   env_file:
  #     - .env
  #   devices:
  #     - /dev/dri:/dev/dri
  #   volumes:
  #     - ./cache:/root/.cache   # Cache hub models and neo_compiler_cache
  #     - ./ollama:/root/.ollama # Cache the ollama models
  #   ports:
  #     - 11434:11434
  #   environment:
  #     - OLLAMA_HOST=0.0.0.0
  #     - DEVICE=Arc
  #     - OLLAMA_INTEL_GPU=true
  #     - OLLAMA_NUM_GPU=999
  #     - ZES_ENABLE_SYSMAN=1
  #     - ONEAPI_DEVICE_SELECTOR=level_zero:0
  #     - TZ=America/Los_Angeles
  #   command: sh -c 'mkdir -p /llm/ollama && cd /llm/ollama && init-ollama && exec ./ollama serve'

  ollama:
    build:
      context: .
      dockerfile: Dockerfile
      target: ollama
    container_name: ollama
    restart: "always"
    env_file:
      - .env
    environment:
      - OLLAMA_HOST=0.0.0.0
      - ONEAPI_DEVICE_SELECTOR=level_zero:0
    devices:
      - /dev/dri:/dev/dri
    ports:
      - 11434:11434 # ollama serve port
    networks:
      - internal
    volumes:
      - ./cache:/root/.cache   # Cache hub models and neo_compiler_cache
      - ./ollama:/root/.ollama # Cache the ollama models
    cap_add: # used for running ze-monitor within container
      - CAP_DAC_READ_SEARCH # Bypass all filesystem read access checks
      - CAP_PERFMON         # Access to perf_events (vs. overloaded CAP_SYS_ADMIN)
      - CAP_SYS_PTRACE      # PTRACE_MODE_READ_REALCREDS ptrace access mode check

  jupyter:
    build:
      context: .
      dockerfile: Dockerfile
      target: jupyter
    #image: jupyter
    container_name: jupyter
    restart: "always"
    env_file:
      - .env
    devices:
      - /dev/dri:/dev/dri
    depends_on:
      - ollama
      - miniircd
    ports:
    - 8888:8888 # Jupyter Notebook
    - 60673:60673 # Gradio
    networks:
      - internal
    volumes:
      - ./jupyter:/opt/jupyter:rw
      - ./cache:/root/.cache

  miniircd:
    build:
      context: .
      dockerfile: Dockerfile
      target: miniircd
    #image: miniircd
    container_name: miniircd
    restart: "no"
    env_file:
      - .env
    ports:
    - 6667:6667 # IRC
    networks:
      - internal
    volumes:
      - ./cache:/root/.cache

  prometheus:
    image: prom/prometheus
    container_name: prometheus
    restart: "always"
    ports:
    - 9090:9090 # Prometheus
    networks:
      - internal
    volumes:
      - ./prometheus.yml:/etc/prometheus/prometheus.yml
      - ./cache/prometheus:/prometheus

  grafana:
    image: grafana/grafana-oss
    container_name: grafana
    restart: "always"
    env_file:
      - .env.grafana
    ports:
    - 3111:3000 # Grafana
    networks:
      - internal
    volumes:
      - ./prometheus.yml:/etc/prometheus/prometheus.yml
      - ./cache/grafana:/var/lib/grafana


#   loki:
#     image: grafana/loki
#     container_name: loki
#     restart: "always"
#     # env_file:
#     #   - .env.grafana
#     ports:
#     - 3211:3100 # Grafana
#     networks:
#       - internal
#     command:
#       - -config.file=/loki-config.yaml
#     volumes:
# #      - ./prometheus.yml:/etc/prometheus/prometheus.yml
#       - ./loki-config.yaml:/loki-config.yaml
#       - ./cache/loki:/loki


networks:
  internal:
    driver: bridge

volumes:
  redis_data:
    driver: local