UnixHTTPConnectionPool(host='localhost', port=None): Read timed out

Использую простой конфиг вида:

  name:
    container_name: name
    build: dockerfile
    command: start app
    working_dir: ./dir
    volumes:
      - ./vol:vol
    env_file:
      - config
    networks:
      - a

В нем развернута tensorflow и простой контроллер на python.

И суть в том, что работает. Но на одной из машин, под ubuntu 20.04 получаю вот эту ошибку:

ERROR: for name  UnixHTTPConnectionPool(host='localhost', port=None): Read timed out. (read timeout=60)
ERROR: An HTTP request took too long to complete. Retry with --verbose to obtain debug information.
If you encounter this issue regularly because of slow network conditions, consider setting COMPOSE_HTTP_TIMEOUT to a higher value (current value: 60).

С --verbose:

compose.parallel.feed_queue: Pending: set()
compose.parallel.feed_queue: Pending: set()
compose.parallel.feed_queue: Pending: set()
compose.parallel.parallel_execute_iter: Failed: ServiceName(project='name', service='name', number=1)
compose.parallel.feed_queue: Pending: set()

ERROR: for name  UnixHTTPConnectionPool(host='localhost', port=None): Read timed out. (read timeout=60)
compose.parallel.parallel_execute_iter: Failed: <Service: name>
compose.parallel.feed_queue: Pending: set()

ERROR: for name  UnixHTTPConnectionPool(host='localhost', port=None): Read timed out. (read timeout=60)
ERROR: compose.cli.errors.log_timeout_error: An HTTP request took too long to complete. Retry with --verbose to obtain debug information.
If you encounter this issue regularly because of slow network conditions, consider setting COMPOSE_HTTP_TIMEOUT to a higher value (current value: 60).

Главное, на более слабых, машинах работает... Даже на 2-ух ядрах. А, на 12-и - нет. Конечно, докер на "проблемной" машине работает.