An incident with a service provider has led me to question my understanding of point-to-point vs. multi-point configurations on my first frame implementation. The provider's tech (whose comprehension I question) insisted that they could not provide me multiple DLCI's to the same physical connection. However, the whole point of subinterfaces is subdividing the same physical connection into separate subnets and DLCI's (in the case of point-to-point). In both p-to-p and multipoint DLCI's have only local significance but if the provider will not provision multiple DLCI's to that one physical interface how can you map them ? Oh, this was a hub and spoke, of course. Every example I see for point-to-point shows remote routers having the same DLCI number with different DLCI's mapped to the core router. Multipoint shows different DLCI's for them all......which has only served to frustrate more. How do you provision these differently with the telco ?