Problems when verifying topology

178
3
05-14-2020 06:03 AM
FabioGrita
New Contributor

Hello, I am really struggling to validate the topology of a dataset. I tried in many ways and with different formats. I also tried to verify a subset of the data (like the one attached), but it takes long time (even days) and at the end I get the error message below. 

I applied 2 basic rules: Must not have gaps (area) and Mist not overlap (area).

I use a personal license of ArcGIS Pro 2.5

Can anybody help?

Many thans

Fabio#

0 Kudos
3 Replies
DuncanHornby
MVP Frequent Contributor

You have multiple issues here:

  1. Having downloaded your sample and built a topology I realised that the test must not have gaps makes no sense as your data has islands.  If they did not have a gap then they would not be an island!
  2. Validation is taking a long time as your data is 1 giant multipart feature, you need to explode that out into single parts, then create your topology on that, then validation is less than a minute for must not have overlaps.
0 Kudos
DanaNolan
Regular Contributor

Topology rules often don't make sense for most of the data. Especially true if someone at a higher level created them for several departments/organizations. They are just there to force you to look at the data and decide whether there is a real error there. I probably have a 500 to 1 exception to real error ration, but that is what exceptions are for.

0 Kudos
DanaNolan
Regular Contributor

I have to work with topologies, not of my design, that generate such errors. There are a huge number of exceptions and overlapping errors and exceptions, but one of the bad areas that triggers the error is less than 100 square meters (out of a million+ acres) and has almost no data in it. So I think the topology engine must be producing bad geometry or just simply can't handle something there. If I turn on Task Manager while these bad areas are being validated, I can see ArcGIS slow down to almost no read/write activity, but it will continue for at least 10 minutes until it fails if I don't cancel validation. 

I mapped the bad areas, which I digitized as a few polygons. Then I validate in sections, avoiding these areas.  This works, but even when I have had to rebuild the topology due to feature name changes, the bad areas are always there. 

0 Kudos