I am trying to do a very basic read/write of an annotation feature class from one file geodatabase to another. In my case, I’m first running a PythonCaller transformer to remove any and all annotation feature classes in the target file geodatabase, then I use a FeatureReader to read the annotation feature class from the source file GDB, and then run a FeatureWriter to write the annotation FC to the target geodatabase with “Insert” and “Create if Needed” options selected for the table, which I assume is creating a new annotation feature class in the target FGDB.

Regardless of minor tweaks I make to both the FeatureReader and FeatureWriter transformers, FME Flow keeps writing these as individual features instead of bulk features. There are ~450k annotation features in the feature class, so running this by writing individual features takes over an hour, which will not work for our scenario.

Interestingly, the first time I ran this in an earlier workspace, FME Flow used the bulk feature setting and the read/write process took ~6 minutes. I’ve been wracking my brain to figure out what I did differently, but it’s like conjuring voodoo to figure out the specific parameters.

Anyone have any ideas of what I’m doing wrong? I wish I could have a copy of the source annotation feature class copied over to the target file geodatabase (in which case the schema’s would already exist and perfectly match), however I have to run a process on the annotation FC in the target FGDB that requires that it be deleted before the workspace is run the following night…

Thanks!