As an OpenShift engineer I would like to go further (if it makes sense) to detect more underlying BM features if the worker node is a BM. For eg. issues Robert and I faced when triaging some customer performance problem were 1) whether the VM master nodes were located on the same BM machine or were they spread across BM machines 2) what does the BM network look like 1Gbps or 10Gbps or more. Basic stuff like this is not known about the underlying infrastructure that makes triaging performance issues a big pain since you have to go back and forth with the infra teams on basic information. Having much of it being exposed through NFD will be helpful.