Schrijven naar Enterprise-database mislukt met fout '-2147216035'

Question

Ik probeer te schrijven naar een van mijn Enterprise-databases zonder succes.<\/P> Het foutbericht is. Het foutnummer van ArcObjects is: '-2147216035'. Het foutbericht van ArcObjects is: Deze geodatabase ondersteunt deze client of bewerking niet.<\/P> De Geodatabase_SDE-schrijver is ingesteld op dynamische schema-definitie, de naam van de feature class is ingesteld op fme_feature_type, en het geometrietype is ingesteld op 'van schema-definitie'... de gebruikelijke instellingen die men zou doen om gegevens met verschillende schema's te schrijven.<\/P> Ik zie dat de naam voor de te schrijven feature class inderdaad de volgende structuur heeft: [fme_feature_type]_[geometrie type].<\/P> Dat verwart me nogal, aangezien de instelling voor de schrijver duidelijk is om alleen fme_feature_type te gebruiken.<\/P> Het andere dat me dwarszit, is het bericht dat de geodatabase deze client of bewerking niet ondersteunt.<\/P> <\/P> Ik gebruik ArcGIS Pro 3.2.1 en de database is versie 10.9.0.2.7 op een MSSQL-server.<\/P>

bharold · Answer

Tom, if you have remaining errors please open a support call.

My guess is the geometry type suffix is an artifact of the reader not knowing the input geometry type, workspace generation will automatically create one (geometry named) feature type per input in those cases.

TomGeo · Answer

The solution for the error of This geodatabase does not support this client or operation. seems to be the database version I was trying to write to.

Since 10.9 versions does not know about datatypes like date_only, and I added an attribute of date_only to the schema the writing part failed.

The suffix of _[geometry type] to the feature class name in the database seems to be a bug.

I am reading the input data from a zipped geopackage, and FME shows the geopackage_type as geopackage_polygon. In my case, the fme_feature_type would be lakes, and the resulting name of the feature class is lakes_polygon.

I tried to circumvent the suffix, fishing with a regular expression for possible parts, separated by an underscore - to no avail.
Since the feature information in the Workbench do show fme_feature_type with a value of lakes, right until the Geodatabase_SDE writer, and even setting the regex in the settings of the writer shows no effect, I assume it is a bug within the writer.

The suffix is also shown when using the Geodatabase_File writer...

After the attempt to write the data into a file geodatabase, and getting the same result as with the enterprise database, I renamed the feature class in the file geodatabase by deleting the suffix.

Then I read the file geodatabase in the same workspace, and used the same dynamic writer to the enterprise database as I used before... this time the name of the resulting feature class in the enterprise database did not receive any suffix and is simply called lakes, as expected.

In the end I used the SchemaScanner, and for good measure I left the regex in the writer settings for the feature class name.

I am using the Data Interoperability Workbench in version 2023.1.1.1, build 23636