Comment 3 for bug 1742401

There is also another problem quite often in this test when "ncat" process don't starts properly:

Captured traceback:
~~~~~~~~~~~~~~~~~~~
    Traceback (most recent call last):
      File "neutron/tests/base.py", line 132, in func
        return f(self, *args, **kwargs)
      File "neutron/tests/fullstack/test_securitygroup.py", line 163, in test_securitygroup
        net_helpers.NetcatTester.TCP)
      File "neutron/tests/fullstack/test_securitygroup.py", line 68, in assert_connection
        self.assertTrue(netcat.test_connectivity())
      File "neutron/tests/common/net_helpers.py", line 509, in test_connectivity
        self.client_process.writeline(testing_string)
      File "neutron/tests/common/net_helpers.py", line 459, in client_process
        self.establish_connection()
      File "neutron/tests/common/net_helpers.py", line 489, in establish_connection
        address=self.address)
      File "neutron/tests/common/net_helpers.py", line 537, in _spawn_nc_in_namespace
        proc = RootHelperProcess(cmd, namespace=namespace)
      File "neutron/tests/common/net_helpers.py", line 288, in __init__
        self._wait_for_child_process()
      File "neutron/tests/common/net_helpers.py", line 321, in _wait_for_child_process
        "in %d seconds" % (self.cmd, timeout)))
      File "neutron/common/utils.py", line 649, in wait_until_true
        raise exception
    RuntimeError: Process ['ncat', u'20.0.0.5', '3333', '-w', '20'] hasn't been spawned in 20 seconds

But I will try to debug it later...