Deployment of instance fails resulting in double network setup

asked 2017-03-17 10:56:33 -0500

Mo Ingolfsson gravatar image


I have a setup with two hypervisors called A and B and Linux bridge Neutron setup. If I spin up an instance that gets assigned to hypervisor A and the spin-up fails for some reason, it gets assigned to hypervisor B. In other words. deployment of instance fails over from compute node A to compute node B. When this happens, it seems that the networking configuration from the original hypervisor (A) does not get cleaned up and the instance ends up with two ip addresses, one of them unusable.

Is this behavior known?

edit retag flag offensive close merge delete