Create auto allocated networks in disabled state
Under particular circumstances, multiple requests to the
auto-allocated-topology extension may lead to the transient
creation of duplicated resources. This is dealt with by the
service plugin code, which cleans them up once the condition
is detected. However the client may accidentally be impacted
and potentially left in error (recoverable on retry).
In order to address this error condition, the logic to
provision the network for any given tenant is tweaked
slightly so that the network is created in disabled state
and re-enabled when it is safe to do so. A Neutron client
should check the network status to see if the network is
ready for use before getting its hands on it.
Closes-bug: #1591766
Change-Id: Ia6ff5ad975673875216eb470080dfc0dcf6b9ab2
diff --git a/neutron/tests/tempest/api/test_auto_allocated_topology.py b/neutron/tests/tempest/api/test_auto_allocated_topology.py
index 1610c85..2c2dfc6 100644
--- a/neutron/tests/tempest/api/test_auto_allocated_topology.py
+++ b/neutron/tests/tempest/api/test_auto_allocated_topology.py
@@ -87,6 +87,8 @@
network_id1 = topology['id']
self.assertIsNotNone(network_id1)
+ network = self.client.show_network(topology['id'])['network']
+ self.assertTrue(network['admin_state_up'])
resources_after1 = self._count_topology_resources()
# One network, two subnets (v4 and v6) and one router
self.assertEqual((1, self.num_subnetpools, 1), resources_after1)