I am currently running a shortest path model but i cannot get any results unless i run a snapping process, even though all of my start and end points line up with nodes in the network. This is fine on small data but i need to run this over a large data set and it is too time consuming. Any ideas on why this would not be working?