r2r/compose.full_with_replicas.yaml

networks:
  r2r-network:
    driver: bridge
    attachable: true
    labels:
      - "com.docker.compose.recreate=always"

volumes:
  hatchet_certs:
    name: ${VOLUME_HATCHET_CERTS:-hatchet_certs}
  hatchet_config:
    name: ${VOLUME_HATCHET_CONFIG:-hatchet_config}
  hatchet_api_key:
    name: ${VOLUME_HATCHET_API_KEY:-hatchet_api_key}
  postgres_data:
    name: ${VOLUME_POSTGRES_DATA:-postgres_data}
  hatchet_rabbitmq_data:
    name: ${VOLUME_HATCHET_RABBITMQ_DATA:-hatchet_rabbitmq_data}
  hatchet_rabbitmq_conf:
    name: ${VOLUME_HATCHET_RABBITMQ_CONF:-hatchet_rabbitmq_conf}
  hatchet_postgres_data:
    name: ${VOLUME_HATCHET_POSTGRES_DATA:-hatchet_postgres_data}

services:
  postgres:
    image: pgvector/pgvector:pg16
    profiles: [postgres]
    environment:
      - POSTGRES_USER=${R2R_POSTGRES_USER:-postgres}
      - POSTGRES_PASSWORD=${R2R_POSTGRES_PASSWORD:-postgres}
      - POSTGRES_HOST=${R2R_POSTGRES_HOST:-postgres}
      - POSTGRES_PORT=${R2R_POSTGRES_PORT:-5432}
      - POSTGRES_MAX_CONNECTIONS=${R2R_POSTGRES_MAX_CONNECTIONS:-1024}
      - PGPORT=${R2R_POSTGRES_PORT:-5432}
    volumes:
      - postgres_data:/var/lib/postgresql/data
    networks:
      - r2r-network
    ports:
      - "${R2R_POSTGRES_PORT:-5432}:${R2R_POSTGRES_PORT:-5432}"
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U ${R2R_POSTGRES_USER:-postgres}"]
      interval: 10s
      timeout: 5s
      retries: 5
    restart: on-failure
    command: >
      postgres
      -c max_connections=${R2R_POSTGRES_MAX_CONNECTIONS:-1024}

  hatchet-postgres:
    image: postgres:latest
    environment:
      POSTGRES_DB: ${HATCHET_POSTGRES_DBNAME:-hatchet}
      POSTGRES_USER: ${HATCHET_POSTGRES_USER:-hatchet_user}
      POSTGRES_PASSWORD: ${HATCHET_POSTGRES_PASSWORD:-hatchet_password}
    volumes:
      - hatchet_postgres_data:/var/lib/postgresql/data
    networks:
      - r2r-network
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U ${HATCHET_POSTGRES_USER:-hatchet_user} -d ${HATCHET_POSTGRES_DBNAME:-hatchet}"]
      interval: 10s
      timeout: 5s
      retries: 5


  hatchet-rabbitmq:
    image: "rabbitmq:3-management"
    hostname: "hatchet-rabbitmq"
    ports:
      - "${R2R_RABBITMQ_PORT:-5673}:5672"
      - "${R2R_RABBITMQ_MGMT_PORT:-15673}:15672"
    environment:
      RABBITMQ_DEFAULT_USER: "user"
      RABBITMQ_DEFAULT_PASS: "password"
    volumes:
      - hatchet_rabbitmq_data:/var/lib/rabbitmq
      - hatchet_rabbitmq_conf:/etc/rabbitmq/rabbitmq.conf
    healthcheck:
      test: ["CMD", "rabbitmqctl", "status"]
      interval: 10s
      timeout: 10s
      retries: 5
    networks:
      - r2r-network

  hatchet-create-db:
    image: postgres:latest
    command: >
      sh -c "
        set -e
        echo 'Waiting for PostgreSQL to be ready...'
        while ! pg_isready -h hatchet-postgres -p 5432 -U ${HATCHET_POSTGRES_USER:-hatchet_user}; do
          sleep 1
        done
        echo 'PostgreSQL is ready, checking if database exists...'
        if ! PGPASSWORD=${HATCHET_POSTGRES_PASSWORD:-hatchet_password} psql -h hatchet-postgres -p 5432 -U ${HATCHET_POSTGRES_USER:-hatchet_user} -lqt | grep -qw ${HATCHET_POSTGRES_DBNAME:-hatchet}; then
          echo 'Database does not exist, creating it...'
          PGPASSWORD=${HATCHET_POSTGRES_PASSWORD:-hatchet_password} createdb -h hatchet-postgres -p 5432 -U ${HATCHET_POSTGRES_USER:-hatchet_user} -w ${HATCHET_POSTGRES_DBNAME:-hatchet}
        else
          echo 'Database already exists, skipping creation.'
        fi
      "
    environment:
      DATABASE_URL: "postgres://${HATCHET_POSTGRES_USER:-hatchet_user}:${HATCHET_POSTGRES_PASSWORD:-hatchet_password}@hatchet-postgres:5432/${HATCHET_POSTGRES_DBNAME:-hatchet}?sslmode=disable"
    networks:
      - r2r-network

  hatchet-migration:
    image: ghcr.io/hatchet-dev/hatchet/hatchet-migrate:latest
    environment:
      DATABASE_URL: "postgres://${HATCHET_POSTGRES_USER:-hatchet_user}:${HATCHET_POSTGRES_PASSWORD:-hatchet_password}@hatchet-postgres:5432/${HATCHET_POSTGRES_DBNAME:-hatchet}?sslmode=disable"
    depends_on:
      hatchet-create-db:
        condition: service_completed_successfully
    networks:
      - r2r-network

  hatchet-setup-config:
    image: ghcr.io/hatchet-dev/hatchet/hatchet-admin:latest
    command: /hatchet/hatchet-admin quickstart --skip certs --generated-config-dir /hatchet/config --overwrite=false
    environment:
      DATABASE_URL: "postgres://${HATCHET_POSTGRES_USER:-hatchet_user}:${HATCHET_POSTGRES_PASSWORD:-hatchet_password}@hatchet-postgres:5432/${HATCHET_POSTGRES_DBNAME:-hatchet}?sslmode=disable"

      HATCHET_CLIENT_GRPC_MAX_RECV_MESSAGE_LENGTH: "${HATCHET_CLIENT_GRPC_MAX_RECV_MESSAGE_LENGTH:-134217728}"
      HATCHET_CLIENT_GRPC_MAX_SEND_MESSAGE_LENGTH: "${HATCHET_CLIENT_GRPC_MAX_SEND_MESSAGE_LENGTH:-134217728}"

      DATABASE_POSTGRES_PORT: "5432"
      DATABASE_POSTGRES_HOST: hatchet-postgres
      DATABASE_POSTGRES_USERNAME: "${HATCHET_POSTGRES_USER:-hatchet_user}"
      DATABASE_POSTGRES_PASSWORD: "${HATCHET_POSTGRES_PASSWORD:-hatchet_password}"
      HATCHET_DATABASE_POSTGRES_DB_NAME: "${HATCHET_POSTGRES_DBNAME:-hatchet}"

      SERVER_TASKQUEUE_RABBITMQ_URL: amqp://user:password@hatchet-rabbitmq:5672/
      SERVER_AUTH_COOKIE_DOMAIN: "http://host.docker.internal:${R2R_HATCHET_DASHBOARD_PORT:-7274}"
      SERVER_URL: "http://host.docker.internal:${R2R_HATCHET_DASHBOARD_PORT:-7274}"
      SERVER_AUTH_COOKIE_INSECURE: "t"
      SERVER_GRPC_BIND_ADDRESS: "0.0.0.0"
      SERVER_GRPC_INSECURE: "t"
      SERVER_GRPC_BROADCAST_ADDRESS: "hatchet-engine:7077"
      SERVER_GRPC_MAX_MSG_SIZE: 134217728
    volumes:
      - hatchet_certs:/hatchet/certs
      - hatchet_config:/hatchet/config
    depends_on:
      hatchet-migration:
        condition: service_completed_successfully
      hatchet-rabbitmq:
        condition: service_healthy
    networks:
      - r2r-network

  hatchet-engine:
    image: ghcr.io/hatchet-dev/hatchet/hatchet-engine:latest
    command: /hatchet/hatchet-engine --config /hatchet/config
    restart: on-failure
    depends_on:
      hatchet-setup-config:
        condition: service_completed_successfully
    ports:
      - "${R2R_HATCHET_ENGINE_PORT:-7077}:7077"
    environment:
      DATABASE_URL: "postgres://${HATCHET_POSTGRES_USER:-hatchet_user}:${HATCHET_POSTGRES_PASSWORD:-hatchet_password}@hatchet-postgres:5432/${HATCHET_POSTGRES_DBNAME:-hatchet}?sslmode=disable"
      SERVER_GRPC_BROADCAST_ADDRESS: "hatchet-engine:7077"
      SERVER_GRPC_BIND_ADDRESS: "0.0.0.0"
      SERVER_GRPC_PORT: "7077"
      SERVER_GRPC_INSECURE: "t"
      SERVER_GRPC_MAX_MSG_SIZE: 134217728
    volumes:
      - hatchet_certs:/hatchet/certs
      - hatchet_config:/hatchet/config
    networks:
      - r2r-network
    healthcheck:
      test: ["CMD", "wget", "-q", "-O", "-", "http://localhost:8733/live"]
      interval: 10s
      timeout: 5s
      retries: 5

  hatchet-dashboard:
    image: ghcr.io/hatchet-dev/hatchet/hatchet-dashboard:latest
    command: sh ./entrypoint.sh --config /hatchet/config
    restart: on-failure
    depends_on:
      hatchet-setup-config:
        condition: service_completed_successfully
    environment:
      DATABASE_URL: "postgres://${HATCHET_POSTGRES_USER:-hatchet_user}:${HATCHET_POSTGRES_PASSWORD:-hatchet_password}@hatchet-postgres:5432/${HATCHET_POSTGRES_DBNAME:-hatchet}?sslmode=disable"
    volumes:
      - hatchet_certs:/hatchet/certs
      - hatchet_config:/hatchet/config
    networks:
      - r2r-network
    ports:
      - "${R2R_HATCHET_DASHBOARD_PORT:-7274}:80"

  setup-token:
    image: ghcr.io/hatchet-dev/hatchet/hatchet-admin:latest
    command: >
      sh -c "
        set -e
        echo 'Starting token creation process...'

        # Attempt to create token and capture both stdout and stderr
        TOKEN_OUTPUT=$$(/hatchet/hatchet-admin token create --config /hatchet/config --tenant-id 707d0855-80ab-4e1f-a156-f1c4546cbf52 2>&1)

        # Extract the token (assuming it's the only part that looks like a JWT)
        TOKEN=$$(echo \"$$TOKEN_OUTPUT\" | grep -Eo 'eyJ[A-Za-z0-9_-]*\.eyJ[A-Za-z0-9_-]*\.[A-Za-z0-9_-]*')

        if [ -z \"$$TOKEN\" ]; then
          echo 'Error: Failed to extract token. Full command output:' >&2
          echo \"$$TOKEN_OUTPUT\" >&2
          exit 1
        fi

        echo \"$$TOKEN\" > /tmp/hatchet_api_key
        echo 'Token created and saved to /tmp/hatchet_api_key'

        # Copy token to final destination
        echo -n \"$$TOKEN\" > /hatchet_api_key/api_key.txt
        echo 'Token copied to /hatchet_api_key/api_key.txt'

        # Verify token was copied correctly
        if [ \"$$(cat /tmp/hatchet_api_key)\" != \"$(cat /hatchet_api_key/api_key.txt)\" ]; then
          echo 'Error: Token copy failed, files do not match' >&2
          echo 'Content of /tmp/hatchet_api_key:'
          cat /tmp/hatchet_api_key
          echo 'Content of /hatchet_api_key/api_key.txt:'
          cat /hatchet_api_key/api_key.txt
          exit 1
        fi

        echo 'Hatchet API key has been saved successfully'
        echo 'Token length:' $${#TOKEN}
        echo 'Token (first 20 chars):' $${TOKEN:0:20}
        echo 'Token structure:' $$(echo $$TOKEN | awk -F. '{print NF-1}') 'parts'
        # Check each part of the token
        for i in 1 2 3; do
          PART=$$(echo $$TOKEN | cut -d. -f$$i)
          echo 'Part' $$i 'length:' $${#PART}
          echo 'Part' $$i 'base64 check:' $$(echo $$PART | base64 -d >/dev/null 2>&1 && echo 'Valid' || echo 'Invalid')
        done
        # Final validation attempt
        if ! echo $$TOKEN | awk -F. '{print $$2}' | base64 -d 2>/dev/null | jq . >/dev/null 2>&1; then
          echo 'Warning: Token payload is not valid JSON when base64 decoded' >&2
        else
          echo 'Token payload appears to be valid JSON'
        fi
      "
    networks:
      - r2r-network
    volumes:
      - hatchet_certs:/hatchet/certs
      - hatchet_config:/hatchet/config
      - hatchet_api_key:/hatchet_api_key
    depends_on:
      hatchet-setup-config:
        condition: service_completed_successfully

  unstructured:
    image: ${UNSTRUCTURED_IMAGE:-ragtoriches/unst-prod}
    networks:
      - r2r-network
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:7275/health"]
      interval: 10s
      timeout: 5s
      retries: 5

  graph_clustering:
    image: ${GRAPH_CLUSTERING_IMAGE:-ragtoriches/cluster-prod}
    ports:
      - "${R2R_GRAPH_CLUSTERING_PORT:-7276}:7276"
    networks:
      - r2r-network
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:7276/health"]
      interval: 10s
      timeout: 5s
      retries: 5

  r2r:
    image: ${R2R_IMAGE:-ragtoriches/prod:latest}
    build:
      context: .
      args:
        PORT: ${R2R_PORT:-7272}
        R2R_PORT: ${R2R_PORT:-7272}
        HOST: ${R2R_HOST:-0.0.0.0}
        R2R_HOST: ${R2R_HOST:-0.0.0.0}
    environment:
      - PYTHONUNBUFFERED=1
      - R2R_PORT=${R2R_PORT:-7272}
      - R2R_HOST=${R2R_HOST:-0.0.0.0}

      # R2R
      - R2R_CONFIG_NAME=${R2R_CONFIG_NAME:-}
      - R2R_CONFIG_PATH=${R2R_CONFIG_PATH:-}
      - R2R_PROJECT_NAME=${R2R_PROJECT_NAME:-r2r_default}

      # Postgres
      - R2R_POSTGRES_USER=${R2R_POSTGRES_USER:-postgres}
      - R2R_POSTGRES_PASSWORD=${R2R_POSTGRES_PASSWORD:-postgres}
      - R2R_POSTGRES_HOST=${R2R_POSTGRES_HOST:-postgres}
      - R2R_POSTGRES_PORT=${R2R_POSTGRES_PORT:-5432}
      - R2R_POSTGRES_DBNAME=${R2R_POSTGRES_DBNAME:-postgres}
      - R2R_POSTGRES_MAX_CONNECTIONS=${R2R_POSTGRES_MAX_CONNECTIONS:-1024}
      - R2R_POSTGRES_STATEMENT_CACHE_SIZE=${R2R_POSTGRES_STATEMENT_CACHE_SIZE:-100}

      # OpenAI
      - OPENAI_API_KEY=${OPENAI_API_KEY:-}
      - OPENAI_API_BASE=${OPENAI_API_BASE:-}

      # Anthropic
      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-}

      # Azure
      - AZURE_API_KEY=${AZURE_API_KEY:-}
      - AZURE_API_BASE=${AZURE_API_BASE:-}
      - AZURE_API_VERSION=${AZURE_API_VERSION:-}

      # Google Vertex AI
      - GOOGLE_APPLICATION_CREDENTIALS=${GOOGLE_APPLICATION_CREDENTIALS:-}
      - VERTEX_PROJECT=${VERTEX_PROJECT:-}
      - VERTEX_LOCATION=${VERTEX_LOCATION:-}

      # AWS Bedrock
      - AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID:-}
      - AWS_SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY:-}
      - AWS_REGION_NAME=${AWS_REGION_NAME:-}

      # Groq
      - GROQ_API_KEY=${GROQ_API_KEY:-}

      # Cohere
      - COHERE_API_KEY=${COHERE_API_KEY:-}

      # Anyscale
      - ANYSCALE_API_KEY=${ANYSCALE_API_KEY:-}

      # Ollama
      - OLLAMA_API_BASE=${OLLAMA_API_BASE:-http://host.docker.internal:11434}

      # Huggingface
      - HUGGINGFACE_API_BASE=${HUGGINGFACE_API_BASE:-http://host.docker.internal:8080}
      - HUGGINGFACE_API_KEY=${HUGGINGFACE_API_KEY}

      # Unstructured
      - UNSTRUCTURED_API_KEY=${UNSTRUCTURED_API_KEY:-}
      - UNSTRUCTURED_API_URL=${UNSTRUCTURED_API_URL:-https://api.unstructured.io/general/v0/general}
      - UNSTRUCTURED_SERVICE_URL=${UNSTRUCTURED_SERVICE_URL:-http://unstructured:7275}
      - UNSTRUCTURED_NUM_WORKERS=${UNSTRUCTURED_NUM_WORKERS:-10}

      # Hatchet
      - HATCHET_CLIENT_TLS_STRATEGY=none
      - HATCHET_CLIENT_GRPC_MAX_RECV_MESSAGE_LENGTH=${HATCHET_CLIENT_GRPC_MAX_RECV_MESSAGE_LENGTH:-134217728}
      - HATCHET_CLIENT_GRPC_MAX_SEND_MESSAGE_LENGTH=${HATCHET_CLIENT_GRPC_MAX_SEND_MESSAGE_LENGTH:-134217728}

      # Graphologic
      - CLUSTERING_SERVICE_URL=http://graph_clustering:7276

    command: >
      sh -c '
        if [ -z "$${HATCHET_CLIENT_TOKEN}" ]; then
          export HATCHET_CLIENT_TOKEN=$$(cat /hatchet_api_key/api_key.txt)
        fi
        exec uvicorn core.main.app_entry:app --host $${R2R_HOST} --port $${R2R_PORT}
      '
    networks:
      - r2r-network
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:${R2R_PORT:-7272}/v3/health"]
      interval: 6s
      timeout: 5s
      retries: 5
    restart: on-failure
    volumes:
      - ${R2R_CONFIG_PATH:-/}:${R2R_CONFIG_PATH:-/app/config}
      - hatchet_api_key:/hatchet_api_key:ro
    extra_hosts:
      - host.docker.internal:host-gateway
    depends_on:
      setup-token:
        condition: service_completed_successfully
      unstructured:
        condition: service_healthy

  r2r-dashboard:
    image: emrgntcmplxty/r2r-dashboard:latest
    environment:
      - NEXT_PUBLIC_R2R_DEPLOYMENT_URL=${R2R_DEPLOYMENT_URL:-http://localhost:7272}
      - NEXT_PUBLIC_HATCHET_DASHBOARD_URL=${HATCHET_DASHBOARD_URL:-http://localhost:${R2R_HATCHET_DASHBOARD_PORT:-7274}}
    networks:
      - r2r-network
    ports:
      - "${R2R_DASHBOARD_PORT:-7273}:3000"

  nginx:
    image: nginx:latest
    ports:
      - "${R2R_NGINX_PORT:-7280}:80"
    volumes:
      - ./nginx.conf:/etc/nginx/nginx.conf:ro
    depends_on:
      r2r:
        condition: service_healthy
    networks:
      - r2r-network
    deploy:
      resources:
        limits:
          cpus: '0.5'
          memory: 512M
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost/health"]
      interval: 10s
      timeout: 5s
      retries: 3