Comment 12 for bug 1873265

Per statement in https://docs.openstack.org/oslo.messaging/latest/reference/transport.html

"""
You may include multiple different network locations separated by commas. The client will connect to any of the available locations and will automatically fail over to another should the connection fail.
"""

So it looks like that oslo.messaging will fail over to another node if it failed to connect to one node. So I assume oslo.messaging _send function always can deliver a rpc call successfully unless all the rabbitmq nodes are disconnected.

Yes, rabbitmq will auto-heal split brain if it is configured appropriately by reboot.

From neutron-ovs-agent log, we can see many error messages as below. So I think rabbitmq has been in abnormal state, otherwise why are there some duplicate message even if it has skipped it.

2020-04-16 08:14:55.260 19871 ERROR oslo.messaging._drivers.impl_rabbit [-] Failed to process message ... skipping it.: DuplicateMessageError: Found duplicate message(7aa9087c4ad047d08af3a1d5ff409834). Skipping it.