docs/doc/source/vnf_integration/use-a-vif-model-when-creating-ports.rst

Using a VIF Model when Creating Ports

For any non Virtio , you must create a port beforehand using the VIF model.

For example, to set up two , the heat template's binding:profile would contain definitions similar to the following.

port_tenant_net_tenant1_avp:
   type: OS::Neutron::Port
   properties:
      name: heat_port_port-tenant-net-tenant1-avp
      network: tenant1-net1
      value_specs: {"binding:profile": {"vif_model":"avp"}}

port_internal_net_tenant1_avp:
   type: OS::Neutron::Port
   properties:
      name: heat_port_port-internal-net-tenant1-avp
      network: internal0-net1
      value_specs: {"binding:profile": {"vif_model":"avp"}}

tenant1_avp:
   type: OS::Nova::Server
   properties:
     name: tenant1-avp4
     flavor: small
     block_device_mapping:
     - {device_name: vda, volume_id: { get_resource: vol_tenant1_avp } }
     networks:
     - {network: tenant1-mgmt-net}
     - {port: { get_resource: port_tenant_net_tenant1_avp } }
     - {port: { get_resource: port_internal_net_tenant1_avp } }

Alternatively, you can use the command line. For example:

~(keystone_admin)$ openstack port create --network ${NETWORKID} --binding-profile vif_model=${VIF_MODEL} ${NAME}

For more information about port related commands, see https://docs.openstack.org/python-openstackclient/train/cli/command-objects/ port.html <https://docs.openstack.org/python-openstackclient/train/cli/command-objects/ port.html>__.