Pipeline crashes when calls to XAPI timeout

Mainly happens when the XNAT server is under heavy load while preprocessing multiple exams. Should probably have the pipeline keep trying until it gets a response (similar to archiving step in the XNAT client).

Traceback:

File "./main.py", line 173, in update_container
    launch_response = launch_next_container(container_dict)
  File "./main.py", line 107, in launch_next_container
    process.update({'command_id': get_command_id(process_name)})
  File "./main.py", line 66, in get_command_id
    with xnat.connect(URL, user=USER, password=PASSWD) as session:
  File "/shared/nonrestricted/cbi-development/crania/venv-donut/lib/python3.6/site-packages/xnat/__init__.py", line 422, in connect
    response = _query_jsession(requests_session, server, debug=debug)
  File "/shared/nonrestricted/cbi-development/crania/venv-donut/lib/python3.6/site-packages/xnat/__init__.py", line 245, in _query_jsession
    raise exceptions.XNATConnectionError('Could not connect to {} (encountered {})'.format(server,
exception_type))
xnat.exceptions.XNATConnectionError: Could not connect to http://xnat.cbi.bnac.net:8080 (encountered ReadTimeout)