When southbound devices are connected to odl cluster via ha proxy.

each client is connected to single odl node.

Some times when the connection flap happens, its node goes missing from operational datastore.

client connects to only one odl controller via ha proxy

1) client disconnects and connects back to same node after some delay
2) client disconnects and connects back to same node immediately
3) client disconnects and connects to another node after some delay
4) client disconnects and connects to another node immediately
5) client disconnects and never connects back

When client disconnects all the odl controllers are trying to cleanup the oper
node.
When client connects the owner odl controller is trying to create the oper node.

When the processing of one odl controller which is trying to cleanup the opernode is
delayed , then we end up client node missing in oper topology.