Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

How does neutron-server distribute tunnel info to neutron-openvswitch-agent in compute/network nodes

Let's say I have 1 control + 1 network + 1 compute node. I am using vxlan with ML2.

Then I boot up another compute node. neutron-openvswitch-agent reads ml2 config file and it will send a message to port 5672 (rabbitmq) in the control node. One of the info contained in that message is the tunnel IP of the new compute node. I am able to trace the code in the compute node.

However, I have no luck in tracking down the code on the control node for it to receive that message and relay that tunneling_ip info to the network node and the first compute node. Could anyone help?

Thanks.

How does neutron-server distribute tunnel info to neutron-openvswitch-agent in compute/network nodes

Let's say I have 1 control + 1 network + 1 compute node. I am using vxlan with ML2.

Then I boot up another compute node. neutron-openvswitch-agent reads ml2 config file and it will send a message to port 5672 (rabbitmq) in the control node. One of the info contained in that message is the tunnel IP of the new compute node. I am able to trace the code in the compute node.

However, I have no luck in tracking down the code on the control node for it to receive that message and relay that tunneling_ip info to the network node and the first compute node. Could anyone help?

Thanks.