The Network Dataset I'm using to do a closest facility analysis keeps giving me the "Error: The network dataset has no edge elements. It may be unbuilt." message. I've ensured that the network is in fact built, and I "edges" show up in my table of contents. I'm unable to identify the actual source of the problem....
I've seen this issue once - the problem gone away after installing SP3 for 10.0.
Otherwise - I'd try several things:
1) Add a network dataset to ArcMap and when prompted do not add any source features. Does it draw correctly and has edges? 2) Create a new file geodatabase with a new feature dataset. Use the Feature Class to Feature Class GP tool to convert the Streets feature class. Build a new network dataset there just to test it would work in a "fresh" environment. 3) Did you upgrade a network dataset from 9.3 to 10 recently? Do you store your network dataset in a geodatabase of 9.3 version or 10?
If the out of memory is during a closest facility solve then how many incidents, facilities do you have and how many are you solving on? And do you have hierarchy on OR a meaningful cutoff in place? For example if the closest facilities should be within 60 minutes, then use that as the distance cutoff.
I've got 787,808 "incidents" and 65,873 facilities. Essentially I have residential locations and commercial/destinations. Closest facility seems like the best option for generating outputs of actual routes from "origins" and "destinations".
I neglected to set a default cut off value before. So it's now set to 5. I don't have a hierarchy in the network. It's just a simple one.
If you try to solve all paths between 787,808 incidents and 65,873 facilities then that will result in about 52 billion paths and the system will run out of memory. Hopefully your 5 minute cutoff will result in a more manageable number.