docker compose – rabbitmq spam logs with `connection` `closing AMQP connection` when celery container has healthcheck

I have docker cluster with celery, rabbitmq and other stuff. And I want to be sure that every service is healthy while cluster is starting. So here is the sample of my docker-compose.yaml, which is related to celery:

  rabbit:
    image: rabbitmq:3.13-management
    container_name: rabbit
    hostname: rabbit
    env_file: .env
    healthcheck:
      test: rabbitmq-diagnostics check_port_connectivity
      interval: 5s
      timeout: 10s
      retries: 10
    volumes:
      - rabbitdata:/var/lib/rabbitmq
    ports:
      - 5672:5672
      - 15672:15672

  celery:
    build:
      context: ./backend
      dockerfile: ./celeryDockerfile
    container_name: celery
    env_file: .env
    healthcheck:
      test: celery -b ${CELERY_BROKER} status || exit 1
      interval: 5s
      timeout: 5s
      retries: 10
    depends_on:
      backend:
        condition: service_healthy
      rabbit:
        condition: service_healthy

  dashboard:
    image: mher/flower
    container_name: dashboard
    env_file: .env
    command: celery flower
    volumes:
      - celery:/app
    ports:
      - 5555:5555
    depends_on:
      backend:
        condition: service_healthy
      rabbit:
        condition: service_healthy
      celery:
        condition: service_healthy

As you can see I have a healthcheck for celery. And if it is on, rabbit spams logs with this:

[info] <0.1422.0> accepting AMQP connection <0.1422.0> (172.18.0.8:56208 -> 172.18.0.3:5672)
[info] <0.1422.0> connection <0.1422.0> (172.18.0.8:56208 -> 172.18.0.3:5672): user 'uid0001' authenticated and granted access to vhost "https://stackoverflow.com/"
[info] <0.1439.0> accepting AMQP connection <0.1439.0> (172.18.0.8:56220 -> 172.18.0.3:5672)
[info] <0.1439.0> connection <0.1439.0> (172.18.0.8:56220 -> 172.18.0.3:5672): user 'uid0001' authenticated and granted access to vhost "https://stackoverflow.com/"
[warning] <0.1422.0> closing AMQP connection <0.1422.0> (172.18.0.8:56208 -> 172.18.0.3:5672, vhost: "https://stackoverflow.com/", user: 'uid0001'):
[warning] <0.1422.0> client unexpectedly closed TCP connection
[warning] <0.1439.0> closing AMQP connection <0.1439.0> (172.18.0.8:56220 -> 172.18.0.3:5672, vhost: "https://stackoverflow.com/", user: 'uid0001'):
[warning] <0.1439.0> client unexpectedly closed TCP connection

As far as I know, healthcheck must do specific number of retries, that you specify with retries: tag. However rabbit continue to spam connect-disconnect info in logs forever, until I stop the cluster. How can I fix that?