Using a separate layer outside of the Geometric Network to set barriers for the trace. The trace is a GP model that will eventually be published to server. The trace is a FIND_CONNECTED on a water network. we would typically use valves to stop the trace but are playing around with other options and scenarios. Placed barriers from a FGDB that would isolate an area and set it as Barriers in the "Trace Geometric Network" GP tool. The Trace GP completely ignores them and try's to trace the entire network. Do I not understand how this is supposed to work or am I doing something wrong? I have verified connectivity and our network is solid. What gives?
Best guess would be that the barriers you are using are not snapped to the line.
if the barrier feature happens to fall on a junction the trace should look at the junction first and may not consider your barrier.
The barrier feature should not be in the GN
Make sure you check the option to Limit results to features stopping the trace.
Thanks for the quick response. The barriers are snapped to the Water Mains and it still ignored them. I also tried putting them on junctions but it did not change anything. Also I have the check box checked for limit results to features stopping the trace.
You got me now. Sounds like your doing everything correct.
Just a shot here. do the barriers have an enabled field. If they are not part of the network it should not make any difference. Just straw grasping.
Simple or complex network? I generally, well not generally, I always work in simple and not complex.
Make a copy of your DB and remake the GM using a different one. I work in Electricity so we have determinate sources which are the only ancillary roles active. Is your network set up with ancillary roles?
It is a complex Network and the barriers are outside of the network with no enable field. I will try making a simple copy of our water network and see if the results are different.
ESRI tech support helped me suss out our problem.
It turns out the barriers weren't the issue. Our network had a large number
of features with the Enabled field set to False. These therefore weren't
participating in the network. Any time a flag was placed on a disabled
feature, the output of the trace was empty. When using the Utility Network
toolbar or the geoprocessing tool, this showed as an empty selection.
However, when the trace is in a model, the output has to go through Collect
Values and Make Feature Layer to be accessible. When the result of the
trace was empty, the Collect Value/Make Feature Layer defaults to passing
on the whole layer. This is therefore an expected behavior. In our case,
enabling all the features (except those that truly shouldn't be, i.e.
abandoned pipes) has fixed our workflow.