model-platform/docker-compose.yml

services:
  migrate:
    build:
      context: .
      dockerfile: backend/Dockerfile
    image: model-development-backend:latest
    logging:
      driver: json-file
      options:
        max-size: "200m"
        max-file: "10"
    restart: "no"
    command:
      - uv
      - run
      - --frozen
      - --no-dev
      - --package
      - backend
      - alembic
      - upgrade
      - head
    environment:
      DATABASE_URL: ${DATABASE_URL:?DATABASE_URL is required}
      INITIAL_ADMIN_PASSWORD: ${INITIAL_ADMIN_PASSWORD:-admin12345}
      UV_OFFLINE: "1"
      UV_NO_SYNC: "1"
      APP_CONFIG_SECRET_KEY: ${APP_CONFIG_SECRET_KEY:?APP_CONFIG_SECRET_KEY is required}

  web:
    build:
      context: .
      dockerfile: frontend/Dockerfile
    image: model-development-web:latest
    logging:
      driver: json-file
      options:
        max-size: "200m"
        max-file: "10"
    restart: unless-stopped
    # Architecture §2.2: this is the only service exposed to the host. The
    # default.conf file is mounted as a template; scripts/nginx-entrypoint.sh
    # parses ${S3_ENDPOINT} and writes the rendered config to
    # /etc/nginx/conf.d/default.conf before exec'ing nginx.
    ports:
      - "${GATEWAY_PORT:-8888}:80"
    environment:
      S3_ENDPOINT: ${S3_ENDPOINT:?S3_ENDPOINT is required}
    depends_on:
      backend:
        condition: service_healthy
      runtime:
        condition: service_healthy
    volumes:
      - ${PWD}/frontend:/usr/share/nginx/html
      - ./default.conf:/etc/nginx/conf.d/default.conf.template:ro
    healthcheck:
      test: [ "CMD-SHELL", "wget -qO- http://127.0.0.1/ >/dev/null" ]
      interval: 10s
      timeout: 3s
      retries: 12
      start_period: 10s

  backend:
    build:
      context: .
      dockerfile: backend/Dockerfile
    image: model-development-backend:latest
    logging:
      driver: json-file
      options:
        max-size: "200m"
        max-file: "10"
    restart: unless-stopped
    # No host port: architecture §2.2 — only Nginx is externally reachable.
    # The previous ``8891:8000`` mapping (P0-1) was removed: the
    # ``/internal/v1/*`` storage control plane is now guarded by a
    # shared ``INTERNAL_SERVICE_TOKEN`` instead of network isolation.
    # No local-FS volume: backend stores everything in S3 (S3_*).
    environment:
      DATABASE_URL: ${DATABASE_URL:?DATABASE_URL is required}
      SERVICE_NAME: model-platform-backend
      SCHEDULE_EVENT_NAMESPACE: ${SCHEDULE_EVENT_NAMESPACE:-model-platform-local}
      JWT_SECRET: ${JWT_SECRET:?JWT_SECRET is required}
      INITIAL_ADMIN_PASSWORD: ${INITIAL_ADMIN_PASSWORD:-admin12345}
      RUNTIME_API_URL: http://runtime:8000
      # P0-1 fix: shared secret required by /internal/v1/* routes.
      INTERNAL_SERVICE_TOKEN: ${INTERNAL_SERVICE_TOKEN:?INTERNAL_SERVICE_TOKEN is required}
      STORAGE_BACKEND: ${STORAGE_BACKEND:-s3}
      LOCAL_STORAGE_BASE_DIR: ${LOCAL_STORAGE_BASE_DIR:-/data}
      # S3_* only matter when STORAGE_BACKEND=s3. Defaults are kept so local
      # mode boots without them; override in .env when switching to s3.
      S3_ENDPOINT: ${S3_ENDPOINT:-http://s3:9000}
      S3_ACCESS_KEY: ${S3_ACCESS_KEY:-}
      S3_SECRET_KEY: ${S3_SECRET_KEY:-}
      S3_WORKSPACE_BUCKET: ${S3_WORKSPACE_BUCKET:-workspace}
      S3_VERSION_BUCKET: ${S3_VERSION_BUCKET:-version}
      S3_RUN_LOG_BUCKET: ${S3_RUN_LOG_BUCKET:-run-log}
      S3_TRASH_BUCKET: ${S3_TRASH_BUCKET:-trash}
      S3_TRASH_RETENTION_DAYS: ${S3_TRASH_RETENTION_DAYS:-30}
      READINESS_TARGETS: ${MYSQL_HOST:?MYSQL_HOST is required}:${MYSQL_PORT:-3306},runtime:8000
      UV_OFFLINE: "1"
      UV_NO_SYNC: "1"
      APP_CONFIG_SECRET_KEY: ${APP_CONFIG_SECRET_KEY:?APP_CONFIG_SECRET_KEY is required}
    depends_on:
      migrate:
        condition: service_completed_successfully
      runtime:
        condition: service_healthy
    volumes:
      - ${PWD}:/app
      - ./data:/data
      - /app/.venv
    healthcheck:
      test: [ "CMD-SHELL", "curl -fsS http://127.0.0.1:8000/health/ready >/dev/null" ]
      interval: 10s
      timeout: 5s
      retries: 18
      start_period: 20s

  runtime:
    build:
      context: .
      dockerfile: runtime/Dockerfile
    image: model-development-runtime:latest
    logging:
      driver: json-file
      options:
        max-size: "200m"
        max-file: "10"
    restart: unless-stopped
    # No host port: architecture §2.2 — only Nginx is externally reachable.
    # The previous ``8892:8000`` mapping (P0-1) was removed: the runtime
    # container is reachable only from the Docker internal network and
    # Nginx-authenticated Jupyter paths.
    cap_add:
      - SYS_ADMIN
    devices:
      - /dev/fuse:/dev/fuse
    security_opt:
      - apparmor:unconfined
    # No host port: architecture §2.2 — only Nginx is externally reachable.
    environment:
      DATABASE_URL: ${DATABASE_URL:?DATABASE_URL is required}
      SERVICE_NAME: runtime-manager
      # P0-1 fix: runtime's /api/v1/jupyter is token-guarded. The same
      # ``INTERNAL_SERVICE_TOKEN`` value backend uses for /internal/v1/*
      # auth — see ``require_internal_service`` in runtime/main.py.
      INTERNAL_SERVICE_TOKEN: ${INTERNAL_SERVICE_TOKEN:?INTERNAL_SERVICE_TOKEN is required}
      STORAGE_BACKEND: ${STORAGE_BACKEND:-s3}
      LOCAL_STORAGE_BASE_DIR: ${LOCAL_STORAGE_BASE_DIR:-/data}
      # WORKSPACES_ROOT defaults to /data/workspace (settings.workspaces_root);
      # in local mode runtime skips the rclone mount and reads directly from
      # ${LOCAL_STORAGE_BASE_DIR}/workspace instead.
      PUBLIC_BASE_URL: http://runtime
      # rclone config only used when STORAGE_BACKEND=s3 (mount skipped in local mode).
      # The remote spec ("s3:<workspace_bucket>") is derived in
      # common.storage.rclone_remote_spec(); no REMOTE_BUCKET env needed.
      RCLONE_CONFIG_S3_TYPE: s3
      RCLONE_CONFIG_S3_PROVIDER: Other
      RCLONE_CONFIG_S3_ACCESS_KEY_ID: ${S3_ACCESS_KEY:-}
      RCLONE_CONFIG_S3_SECRET_ACCESS_KEY: ${S3_SECRET_KEY:-}
      RCLONE_CONFIG_S3_ENDPOINT: ${S3_ENDPOINT:-http://s3:9000}
      RCLONE_CONFIG_S3_ENV_AUTH: "false"
      RCLONE_CONFIG_S3_FORCE_PATH_STYLE: "true"
      RCLONE_CONFIG_S3_REGION: other
      UV_OFFLINE: "1"
      UV_NO_SYNC: "1"
      APP_CONFIG_SECRET_KEY: ${APP_CONFIG_SECRET_KEY:?APP_CONFIG_SECRET_KEY is required}
    volumes:
      - ${PWD}:/app
      - ./data:/data
      - /app/.venv
    healthcheck:
      # Two-mode: s3 needs rclone FUSE mount; local skips rclone and uses a
      # bind mount instead. In s3 mode the grep succeeds (mount must be up or
      # startup would have crashed). In local mode the grep fails (no fuse.rclone)
      # so the || branch runs: just verify /data/workspace is a directory.
      test: [ "CMD-SHELL", "curl -fsS http://127.0.0.1:8000/api/v1/health >/dev/null" ]
      interval: 10s
      timeout: 5s
      retries: 18
      start_period: 30s

  schedule:
    build:
      context: .
      dockerfile: schedule/Dockerfile
    image: model-development-schedule:latest
    logging:
      driver: json-file
      options:
        max-size: "200m"
        max-file: "10"
    restart: unless-stopped
    # No host port: architecture §2.2 — only Nginx is externally reachable.
    # No local-FS volume: schedule executes nodes via tempfile.TemporaryDirectory
    # under Python's default temp dir (cleaned per-run); artifacts live in S3.
    environment:
      DATABASE_URL: ${DATABASE_URL:?DATABASE_URL is required}
      SERVICE_NAME: schedule-executor
      SCHEDULE_EVENT_NAMESPACE: ${SCHEDULE_EVENT_NAMESPACE:-model-platform-local}
      BACKEND_API_URL: http://backend:8000
      # P0-1 fix: must match the backend's INTERNAL_SERVICE_TOKEN exactly.
      INTERNAL_SERVICE_TOKEN: ${INTERNAL_SERVICE_TOKEN:?INTERNAL_SERVICE_TOKEN is required}
      STORAGE_BACKEND: ${STORAGE_BACKEND:-s3}
      LOCAL_STORAGE_BASE_DIR: ${LOCAL_STORAGE_BASE_DIR:-/data}
      S3_ENDPOINT: ${S3_ENDPOINT:-http://s3:9000}
      S3_ACCESS_KEY: ${S3_ACCESS_KEY:-}
      S3_SECRET_KEY: ${S3_SECRET_KEY:-}
      S3_WORKSPACE_BUCKET: ${S3_WORKSPACE_BUCKET:-workspace}
      S3_VERSION_BUCKET: ${S3_VERSION_BUCKET:-version}
      S3_RUN_LOG_BUCKET: ${S3_RUN_LOG_BUCKET:-run-log}
      # Backend's health endpoint already verifies MySQL; the scheduler only
      # needs MySQL and Backend to be ready in either local or S3 mode.
      READINESS_TARGETS: ${MYSQL_HOST:?MYSQL_HOST is required}:${MYSQL_PORT:-3306},backend:8000
      UV_OFFLINE: "1"
      UV_NO_SYNC: "1"
      APP_CONFIG_SECRET_KEY: ${APP_CONFIG_SECRET_KEY:?APP_CONFIG_SECRET_KEY is required}
    depends_on:
      backend:
        condition: service_healthy
    volumes:
      - ${PWD}:/app
      - ./data:/data
      - /app/.venv
    healthcheck:
      test: [ "CMD-SHELL", "curl -fsS http://127.0.0.1:8000/health/ready >/dev/null" ]
      interval: 10s
      timeout: 5s
      retries: 18
      start_period: 20s