When uploading a table as part of a feature service one of the table columns is missing and values are shifted into the wrong column. I am attempting to upload a layer with a related table (1:many relationship, based on this blog: Related Tables – Exploring New Ways to use Collector for ArcGIS | ArcGIS Blog ). The uploaded result is a faulty table and the layer does not display its related content.
The table has the following structure in its correct offline form:
PARCELID LINK DOCUMENT TYPE
9001 >linky name
9002 >linky1 name1
Once uploaded the table looks as such:
PARCELID LINK DOCUMENT TYPE
>linky name
>linky1 name1
did you link parcel id to parcel id ? or link to link? you didn't specify
parcel id to parcel id.
The problem with the table also occurred when I uploaded it to a service without the relate.
Do your field names contain reserved words or spaces? See: FAQ: What characters should not be used in ArcGIS for field names and table names? and FAQ: What are the reserved words for Esri's file geodatabase?
Changed "DOCUMNENT TYPE" to "DOCTYPE" just to be sure and the problem persists.
I'm thinking this is a bug with how the publishing process works for standalone data tables.
My next suggestion would be to look at the JSON file and check if the field definitions carried across correctly. For example, is the link field remaining a long/integer or is it becoming text? What type is the doctype field; text or other?
The linked field, a string is remaining a string. Opening the JSON file associated with the uploaded file reveals that all the components in the table are alive and healthy.