VXLAN tunnels on br-tun by Mgmt and Data Interfaces [closed]
Hi all,
we have a 1 controller node + 2 computer nodes, one of them is the network node. At the computer nodes, we had only 1 mgmt interface with addressing 10.50.0.23X and the br-tun vxlan tunnel was formed over this interfaces.
We just add a new interface on both computer nodes with 10.13.111.12X IP address. So, we have changed the br-tun vxlan local_ip to the new ones, but the br-tun vxlan tunnels is pointing also to the old interfaces.
Here you have the ovs .ini config file of compute-02 node (all the outputs are from this node):
[15:46:46][root@ostack-02 ~]$ cat /etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini | grep -v -P "#|^$"
[ovs]
enable_tunneling = True
tunnel_type = vxlan
tunnel_id_ranges = 1:1000
integration_bridge = br-int
tunnel_bridge = br-tun
local_ip = 10.13.111.122
[agent]
polling_interval = 2
tunnel_types =vxlan
vxlan_udp_port =4789
l2_population = False
arp_responder = False
enable_distributed_routing = False
[securitygroup]
firewall_driver = neutron.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver
[15:47:10][root@ostack-02 ~]$
On the following output you can see a sample of the ifconfig with eth2.50 used for mgmt and eth4.917 used for data:
eth2.50: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 10.50.0.232 netmask 255.255.255.0 broadcast 10.50.0.255
ether f8:bc:12:b2:d8:90 txqueuelen 0 (Ethernet)
eth4.917: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 10.13.111.122 netmask 255.255.255.0 broadcast 10.13.111.255
ether f8:bc:12:b2:de:42 txqueuelen 0 (Ethernet)
Here you can see how the br-tun looks like, with 3 different vxlan tunnels:
[15:25:07][root@ostack-02 ~]$ ovs-vsctl show
d9e769f8-103e-4ccc-9714-025a1181fd0e
Bridge br-tun
Port "vxlan-0a0d6f79"
Interface "vxlan-0a0d6f79"
type: vxlan
options: {df_default="true", in_key=flow, local_ip="10.13.111.122", out_key=flow, remote_ip="10.13.111.121"}
Port br-tun
Interface br-tun
type: internal
Port "vxlan-0a3200e8"
Interface "vxlan-0a3200e8"
type: vxlan
options: {df_default="true", in_key=flow, local_ip="10.13.111.122", out_key=flow, remote_ip="10.50.0.232"}
Port patch-int
Interface patch-int
type: patch
options: {peer=patch-tun}
Port "vxlan-0a3200e7"
Interface "vxlan-0a3200e7"
type: vxlan
options: {df_default="true", in_key=flow, local_ip="10.13.111.122", out_key=flow, remote_ip="10.50.0.231"}
Bridge br-int
fail_mode: secure
Port "qvo6307173b-dd"
tag: 1
Interface "qvo6307173b-dd"
Port br-int
Interface br-int
type: internal
Port patch-tun
Interface patch-tun
type: patch
options: {peer=patch-int}
Bridge br-ex
Port br-ex
Interface br-ex
type: internal
ovs_version: "2.1.3"
[15:25:07][root@ostack-02 ~]$
We've set up two VMs on different nodes and they can ping each other, so the the br-tun and vxlan communication between the compute hosts is working.
So my question is:
- As far as I understand, there should be only one vxlan interface under the br-tun, that's the one with the eth4.917 data interface: "local_ip="10.13.111.122", remote_ip="10.13.111.121". Am I right?
- If the above is right, why and what is the purpose of the other two interfaces? Is it due to a misconfiguration?
We've tried to ...
see https://bugs.launchpad.net/neutron/+b...