在OpenStack中，随着网络环境费复杂增长，有时候会遇到dhcp agent port bonding failed，导致agent无法正常提供服务，我们可以简单的用下面的方法来处理
1、查找down掉的port
root@node-32:~# neutron port-list -c id -c tenant_id --status DOWN --device_owner network:dhcp -c device_owner -c device_id -c fixed_ips
+--------------------------------------+----------------------------------+--------------+-------------------------------------------------------------------------------+------------------------------------------------------------------------------------+
| id | tenant_id | device_owner | device_id | fixed_ips |
+--------------------------------------+----------------------------------+--------------+-------------------------------------------------------------------------------+------------------------------------------------------------------------------------+
| fa172acf-8900-487a-9aaf-ff4c12061b2e | d0a33ff6b4da4714a3cea18ddb6dc859 | network:dhcp | dhcp0f308520-710f-57c8-af30-18395065ae2a-68861e75-22c8-4575-bcc3-29b19a8cb189 | {"subnet_id": "ebb540a8-8284-4ec8-89a2-bf180d706a0d", "ip_address": "10.xxx.64.1"} |
+--------------------------------------+----------------------------------+--------------+-------------------------------------------------------------------------------+------------------------------------------------------------------------------------+
2、delet 掉端口，agent服务会重新绑定一个新的端口
root@node-32:~# neutron port-delete fa172acf-8900-487a-9aaf-ff4c12061b2e