NSX|V3: don't fail on already deleted network/port

If under lolad a network deletion fails for the internal metadata
then swallow the DB exception and continue. No need to fail the
neutron opertaion here.

If there are orphan metadata networks afterwards then they can be
cleaned manually.

Change-Id: Ib4f657bed84a2c84363597f861c39cb7fbde7b18

vmware_nsx/dhcp_meta/rpc.py

@@ -18,6 +18,7 @@ from eventlet import greenthread
 from neutron_lib import constants as const
 from neutron_lib import exceptions as ntn_exc
 from oslo_config import cfg
+from oslo_db import exception as db_exc
 from oslo_log import log as logging
 
 from neutron.api.rpc.agentnotifiers import dhcp_rpc_agent_api
@@ -233,6 +234,8 @@ def _destroy_metadata_access_network(plugin, context, router_id, ports):
         # must re-add the router interface
         plugin.add_router_interface(context, router_id,
                                     {'subnet_id': meta_sub_id})
+    except db_exc.DBReferenceError as e:
+        LOG.debug("Unable to delete network %s. Reason: %s", meta_net_id, e)
     # Tell to stop the metadata agent proxy
     _notify_rpc_agent(
         context, {'network': {'id': meta_net_id}}, 'network.delete.end')