How do I bulk update a branch versioned feature class without triggering editor tracking attributes

1843
12
01-10-2023 02:09 PM
Labels (3)
DrewDowling
Occasional Contributor III

We have a feature class that is part of a parcel fabric. The feature class is branch versioned and published through a feature service.

I need to do a bulk update on all features and don't wish to trigger the editor tracking attributes. In ArcMap I was able to do this by doing a field calc outside of an edit session. Is there an equivalent way of doing this in Pro with branch versioned data?

 

0 Kudos
12 Replies
LanceCole
MVP Regular Contributor

Did you try and disable editor tracking before performing the bulk update then enable tracking again?  

Enable or disable editor tracking 

DrewDowling
Occasional Contributor III

I really don't want to go down the path of disabling editor tracking or unregistering the versioning. I went through this before back in the arcmap days and I found it very messy, especially when the feature class was part of a complex dataset like Fabric or Network.

 

0 Kudos
LanceCole
MVP Regular Contributor

I can understand your reservation.

I had recommended using an Edit Session in python to perform your bulk update.  We use this regularly to update attributes in our versioned databases without modifying the editor tracking attributes.  However, I noted you are using a "branch version," which must be edited via a feature service.  Not sure how that plays into editing data without updating the editor tracking attributes.  I have been avoiding going down that path.

DrewDowling
Occasional Contributor III

I played around with your suggestion of using different combinations of cursors and  field calculate both inside and outside an edit session and every time the update worked the editor tracking also got triggered. I was using direct database connects and not the feature service.

So looks like I'll have to figure out another method, probably use the non esri approved direct SQL.

Thanks for your help Lance.

 

0 Kudos
LanceCole
MVP Regular Contributor

Yea, I edited my prior post just after submitting the reply.  I had not noted you are using a branch version that requires editing via a feature service.  Not sure if you can get around not updating the editing data in that situation.  I do not have a test environment on an enterprise deployment to experiment with.  I get myself in trouble experimenting with the production systems.

RiccardoKlinger
Occasional Contributor

Hi Drew,

we are facing the same problem right now as we do mass updates on features quite regularily. Have you found an approach that is in line with the recommondations from Esri and still avoids the duplication of features in the database? Or did you face anyreal problems with direct sql updates?

best, Riccardo

0 Kudos
Andrew_McClary
New Contributor

Hi Drew and Riccardo,

We're facing the same issue using branch versioning. Have you found any good solutions to this or issues with the direct SQL updates? Also, we're finding editing using branch versioning incredibly slow (12 hours to update 40,000 rows). Are other people encountering this problem as well or is it something specific to our system?

Thanks,

Andrew

0 Kudos
DrewDowling
Occasional Contributor III

@Andrew_McClary, @RiccardoKlinger 

I never figured out a way to do this with the Pro client.

I was able to do it using SQL. The only issue I noticed was that the version difference comparison tool stopped working. This is the tool in the Pro versioning ribbon that lets you see differences between versions. I don't know for sure if the SQL edit broke the tool, it is only a guess.

If I created a new feature service with a different name then the version difference tool worked again.

 

@RiccardoKlingerI get slow performance on a large number of edits as well using feature services. In my case it was the topology. I had better success in bulk updates by disabling topology first and then doing the edit.

 

ArizonaGIS
New Contributor III

@DrewDowling, @Andrew_McClary, @RiccardoKlinger & @LanceCole did you figure out a solution? I am attempting to do edits on a feature service that is part of a branch versioning dataset. I am running a field calculation of 200K records to update one field to populate the values to Null, and the calculation is incredibly slow, almost to the point it doesn't seem to be writing to the service.

0 Kudos