If nodes events fail to be handled they are added to different sync maps, but are never retried unless another event comes along later. We should actively retry to add the node, using logic similar to the pod and network policy controllers.

https://github.com/ovn-org/ovn-kubernetes/pull/2870