[openstack-dev] [Neutron]The components timing problem

Sławomir Kapłoński slawek at kaplonski.pl
Tue Jan 23 09:27:52 UTC 2018


Hi,

If both ovs agent and neutron-server reconnect to rabbitmq then it should report state properly again IMO.
Can You maybe send more details about Your issue? What OpenStack version You are running, exact stack trace of exception which You get and so on.

— 
Best regards
Slawek Kaplonski
slawek at kaplonski.pl



> Wiadomość napisana przez Frank Wang <wangpeihuixyz at 126.com> w dniu 23.01.2018, o godz. 10:08:
> 
> Hi All,
> 
>     I'm really newbie about OpenStack Neutron, Please correct me if I say something wrong. There was a question I'd like to consult.  AMQP is the messaging bus between neutron-server and *agents. we usually use rabbitmq as the back-end of messaging bus.  The problem I encountered is the ovs agent raise an exception while reporting its own state to the server.  Here is my guess, If I restart the controller node, what if the rabbitmq start early than neutron-server. I mean the ovs agent always trying to connect to the rabbitmq. It will report state to the server through RPC once the connection established. if the server is not ready at this time. Does it cause the agent exception?  Any suggestion would be greatly appreciated!
> 
> 
> Thanks,
> Frank
> 
> 
>  
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev




More information about the OpenStack-dev mailing list