ai-servers/docker-compose.yml

services:
  # ── Cache for SearXNG ──
  valkey:
    image: valkey/valkey:8-alpine
    command: valkey-server --save 30 1 --loglevel warning
    volumes:
      - valkey-data:/data
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "valkey-cli", "ping"]
      interval: 10s
      timeout: 3s
      retries: 3

  # ── Private search engine ──
  searxng:
    image: searxng/searxng:latest
    volumes:
      - ./searxng:/etc/searxng:rw
    environment:
      - SEARXNG_SECRET_KEY=${SEARXNG_SECRET_KEY}
    ports:
      - "0.0.0.0:8080:8080"
    depends_on:
      valkey:
        condition: service_healthy
    restart: unless-stopped

  # ── Vector database ──
  chromadb:
    image: chromadb/chroma:latest
    volumes:
      - chromadb-data:/chroma/chroma
    ports:
      - "0.0.0.0:8000:8000"
    environment:
      - IS_PERSISTENT=TRUE
      - ANONYMIZED_TELEMETRY=FALSE
    restart: unless-stopped

  # # ── Database for LiteLLM (DEPRECATED — kept for rollback) ──
  # litellm-db:
  #   image: postgres:16-alpine
  #   volumes:
  #     - litellm-db-data:/var/lib/postgresql/data
  #   environment:
  #     - POSTGRES_DB=litellm
  #     - POSTGRES_USER=litellm
  #     - POSTGRES_PASSWORD=${LITELLM_DB_PASSWORD}
  #   restart: unless-stopped
  #   healthcheck:
  #     test: ["CMD-SHELL", "pg_isready -U litellm"]
  #     interval: 10s
  #     timeout: 3s
  #     retries: 3

  # # ── LLM API proxy (DEPRECATED — replaced by new-api) ──
  # litellm:
  #   image: ghcr.io/berriai/litellm:main-latest
  #   command: ["--config", "/app/config.yaml", "--port", "4000"]
  #   volumes:
  #     - ./litellm/config.yaml:/app/config.yaml:ro
  #   ports:
  #     - "0.0.0.0:4000:4000"
  #   environment:
  #     - LITELLM_MASTER_KEY=${LITELLM_MASTER_KEY}
  #     - DATABASE_URL=postgresql://litellm:${LITELLM_DB_PASSWORD}@litellm-db:5432/litellm
  #     - OPENROUTER_API_KEY=${OPENROUTER_API_KEY}
  #     - SILICONFLOW_API_KEY=${SILICONFLOW_API_KEY}
  #     - DEEPINFRA_API_KEY=${DEEPINFRA_API_KEY}
  #     - GROQ_API_KEY=${GROQ_API_KEY}
  #     - CEREBRAS_API_KEY=${CEREBRAS_API_KEY}
  #   depends_on:
  #     litellm-db:
  #       condition: service_healthy
  #   restart: unless-stopped
  #   healthcheck:
  #     test: ["CMD", "python", "-c", "import urllib.request; urllib.request.urlopen('http://localhost:4000/health/liveliness')"]
  #     interval: 15s
  #     timeout: 5s
  #     retries: 5
  #     start_period: 30s

  # ── LLM API proxy (new-api) ──
  new-api:
    image: calciumion/new-api:latest
    ports:
      - "0.0.0.0:4000:3000"
    volumes:
      - new-api-data:/data
    environment:
      - SQL_DSN=
      - TZ=UTC
      - ENABLE_METRIC=true
      - LANG=en_US.UTF-8
      - INITIAL_ROOT_ACCESS_TOKEN=${NEW_API_ACCESS_TOKEN}
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "wget", "-q", "-O", "/dev/null", "http://localhost:3000/"]
      interval: 15s
      timeout: 5s
      retries: 5
      start_period: 10s

  # ── Chat UI ──
  open-webui:
    image: ghcr.io/open-webui/open-webui:main
    volumes:
      - open-webui-data:/app/backend/data
    ports:
      - "0.0.0.0:3000:8080"
    environment:
      - OLLAMA_BASE_URL=
      - OPENAI_API_BASE_URL=http://new-api:3000/v1
      - OPENAI_API_KEY=${OPENWEBUI_API_KEY}
      - ENABLE_RAG_WEB_SEARCH=true
      - RAG_WEB_SEARCH_ENGINE=searxng
      - SEARXNG_QUERY_URL=http://searxng:8080/search?q=<query>&format=json
      - CHROMA_HTTP_HOST=chromadb
      - CHROMA_HTTP_PORT=8000
      - WEBUI_AUTH=true
    depends_on:
      new-api:
        condition: service_healthy
    restart: unless-stopped

  # ── Cloudflare Tunnel (public HTTPS) ──
  cloudflared:
    image: cloudflare/cloudflared:latest
    command: tunnel run
    environment:
      - TUNNEL_TOKEN=${CLOUDFLARE_TUNNEL_TOKEN}
    restart: unless-stopped

  # ── Tailscale (private VPN access) ──
  tailscale:
    image: tailscale/tailscale:latest
    hostname: ai-proxy
    volumes:
      - tailscale-state:/var/lib/tailscale
      - /dev/net/tun:/dev/net/tun
    cap_add:
      - NET_ADMIN
      - SYS_MODULE
    environment:
      - TS_AUTHKEY=${TS_AUTHKEY}
      - TS_STATE_DIR=/var/lib/tailscale
      - TS_USERSPACE=false
    restart: unless-stopped
    network_mode: host

  # ═══════════════════════════════════════════════
  # Monitoring stack
  # ═══════════════════════════════════════════════

  # ── Metrics store (Prometheus-compatible) ──
  victoriametrics:
    image: victoriametrics/victoria-metrics:latest
    volumes:
      - victoriametrics-data:/victoria-metrics-data
      - ./monitoring/prometheus.yml:/etc/prometheus/prometheus.yml:ro
    command:
      - "-promscrape.config=/etc/prometheus/prometheus.yml"
      - "-retentionPeriod=90d"
      - "-storageDataPath=/victoria-metrics-data"
    ports:
      - "127.0.0.1:8428:8428"
    restart: unless-stopped

  # ── Dashboards ──
  grafana:
    image: grafana/grafana:latest
    volumes:
      - grafana-data:/var/lib/grafana
    ports:
      - "0.0.0.0:3001:3000"
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_ADMIN_PASSWORD}
      - GF_USERS_ALLOW_SIGN_UP=false
    depends_on:
      - victoriametrics
    restart: unless-stopped

  # ── Host system metrics ──
  node-exporter:
    image: prom/node-exporter:latest
    pid: host
    volumes:
      - /proc:/host/proc:ro
      - /sys:/host/sys:ro
      - /:/rootfs:ro
    command:
      - "--path.procfs=/host/proc"
      - "--path.sysfs=/host/sys"
      - "--path.rootfs=/rootfs"
      - "--collector.filesystem.mount-points-exclude=^/(sys|proc|dev|host|etc)($$|/)"
    restart: unless-stopped

  # ── Valkey/Redis metrics ──
  redis-exporter:
    image: oliver006/redis_exporter:latest
    environment:
      - REDIS_ADDR=redis://valkey:6379
    depends_on:
      valkey:
        condition: service_healthy
    restart: unless-stopped

volumes:
  valkey-data:
  chromadb-data:
  litellm-db-data:
  new-api-data:
  open-webui-data:
  tailscale-state:
  victoriametrics-data:
  grafana-data: