Replace Polygon geometry in SDE Database using FME

Question

I'm still trying to wrap my head around using FME and encountering another issue.

This is my scenario: I have a master SDE database that i am trying to update using the data from a GDB file. I was able to successfully update the SDE database on the Points data but when i tried to do the same process on Polygon's, the update breaks my SDE data.

FME Translation Setup:

• I have 2 readers (GDB file reader and SDE DB reader)
• I created 3 writers (INSERT, DELETE, UPDATE) *there could be a better implementation
• I am just reusing these readers and writers and created an update process for each feature inside the GDB file.

Process: enter image description here

1. Read through the SDE and GDB data.
2. Run CRC on each dataset.
3. Merge resulting datasets using FeatureMerger
   • Type: Attributes and Geometry
4. Run GeometryCoercer and 2DForcer.
5. Tester, check if both CRC fields are the same.
   • If yes, do nothing
   • If no, update the row.

The translation succeeds but when i try to view the feature in ArcCatalog, it gives me this error: ArcCatalogError

I might just be doing/setting something stupid that is causing this issue.

UPDATE: Also tested the updated geometries before the insert using the Data Inspector and everything seems to be fine. I think my issue is somewhere in the WRITER.

• Mapperz: When you do an INSERT you are required to find the Next_RowID takes a table that is registered with the geodatabase as an input parameter and returns the next value for the ObjectID (RowID) field. resources.arcgis.com/en/help/main/10.1/index.html#//…
• Juno: hi Mapperz, i only tested the update first and disabled the INSERT and DELETE. Also, the data in the GDB files already has the correct value for the ObjectId, so i don't think i should worry about that for now.

Answer 1

I think i just figured out why i was encountering this issue. I was doing something stupid and used a Microsoft Spatial format for my reader and writer. When i used the ESRI geodatabase format instead, the update works like a charm.