[neutron][ovn-bgp-agent] ovs bridges down causes error when configuring route table
Hi, I am trying to use the OVN BGP agent (nb_ovn_bgp_driver with the exposing method underlay) and I would like to share a difficulty that I found, it might be related to the specific deployment that I have using Kayobe. I find that by default my openvswitch bridges appear to be down when I list them with "ip link". From what I saw in some forums this is a normal behaviour. The thing is that the OVN BGP agent uses ip link, for example here [1], to add route to a table link to these bridges and if there are down, this raise an error(pyroute2.netlink.exceptions.NetlinkError: (100, 'Network is down')). When I set the bridge state UP with "ip link" it solves the issue but this solution is not reboot persistent. Is it really a normal behaviour to have the openvswitch bridges down or is it a problem that I need to investigate on my configuration ? If this is a normal behaviour, anyone found a workaround that could be resilient in production (that is reboot persistent) ? Do not hesitate if you need more information on my configuration ! [1]: https://opendev.org/openstack/ovn-bgp-agent/src/branch/master/ovn_bgp_agent/...
Greetings, yes, that the ovs bridge is down is normal behaviour. In our deployment the ovn bgp agent systemd unit has a ExecStartPre condition to bring up the interface before the agent starts. Also the ovn bgp agent got a patch to do this automatically: https://github.com/openstack/ovn-bgp-agent/commit/fa344056c8c98b537cf2224340... This patch is in release >= 5.0.0 Best Regards Max ________________________________ From: anas.jouhdy@gmail.com <anas.jouhdy@gmail.com> Sent: 14 January 2026 16:34 To: openstack-discuss@lists.openstack.org <openstack-discuss@lists.openstack.org> Subject: [neutron][ovn-bgp-agent] ovs bridges down causes error when configuring route table Warning: This email is from an unusual correspondent. Warning: Make sure this is someone you trust. Hi, I am trying to use the OVN BGP agent (nb_ovn_bgp_driver with the exposing method underlay) and I would like to share a difficulty that I found, it might be related to the specific deployment that I have using Kayobe. I find that by default my openvswitch bridges appear to be down when I list them with "ip link". From what I saw in some forums this is a normal behaviour. The thing is that the OVN BGP agent uses ip link, for example here [1], to add route to a table link to these bridges and if there are down, this raise an error(pyroute2.netlink.exceptions.NetlinkError: (100, 'Network is down')). When I set the bridge state UP with "ip link" it solves the issue but this solution is not reboot persistent. Is it really a normal behaviour to have the openvswitch bridges down or is it a problem that I need to investigate on my configuration ? If this is a normal behaviour, anyone found a workaround that could be resilient in production (that is reboot persistent) ? Do not hesitate if you need more information on my configuration ! [1]: https://urlsand.esvalabs.com/?u=https%3A%2F%2Fopendev.org%2Fopenstack%2Fovn-bgp-agent%2Fsrc%2Fbranch%2Fmaster%2Fovn_bgp_agent%2Futils%2Flinux_net.py%23L293&e=50525cf7&h=ac6457ba&f=y&p=y -- This message has been checked by Libraesva ESG and is found to be clean. Follow this link to report as spam/bad: https://mx10.wiit.cloud/action/4drr0J69HpzTgWm/report-as-bad Follow this link to blocklist sender: https://mx10.wiit.cloud/action/4drr0J69HpzTgWm/blocklist
Hi, Thank you very much for the response, it really helps. I did not know that it was patch! I will check if I can use it ! Best regards,
participants (3)
-
Anas Jouhdy
-
anas.jouhdy＠gmail.com
-
Maximilian Stinsky-Damke