Description of problem:

There is a dispatcher script deployed on OCP nodes to ensure that OVNK keeps working after a NetworkManager restart. However, it assumes the profiles will be named the way configure-ovs does, and NMState uses different profile names by default. We should update the script to support the NMState names as well so NM restarts work correctly.

Version-Release number of selected component (if applicable):

We'll want to fix this back to 4.16 when the NMState feature first became available.

How reproducible:

Always

Steps to Reproduce:

    1. Deploy cluster with NMState managing br-ex
    2. Restart NetworkManager on a node
    3.
    

Actual results:

Pod connectivity is broken    

Expected results:

    Connectivity continues working

Additional info: