Optimizing nmstate object handling on OpenShift 4

Issue

• nncp application takes a long time
• nncp objects become degraded when one node in your cluster fails to apply configuration changes
• nncp objects are degraded when an interface they are dependent upon doesn't come up first (also defined by nncp object).
• node will take nncp configuration just fine but then after restart will fail to return to READY status/networking degraded on subsequent boots.