Here is the python code I am using for the CreateReplica geoprocessing tool:
arcpy.CreateReplica_management("'Database Connections\\MySDEConnection.sde\\DB.OWNER.Table1';'Database Connections\\MySDEConnection.sde\\DB.OWNER.Table2'", "CHECK_OUT", "C:\Temp\MyFileGDB.gdb", "MyCheckoutVersion", "FULL", "CHILD_DATA_SENDER", "ALL_ROWS", "DO_NOT_REUSE", "GET_RELATED", "", "DO_NOT_USE_ARCHIVING")
I changed the name of the connection and of the table names to make them generic for this example.
As I said in my first post, the parameter that I have set to "ALL_ROWS" is supposed to check out all rows for both feature classes and tables. What is happening for me is that it only checks out the schema for the tables -- no data/no records. I would think that ALL_ROWS should be the default, but it isn't. The default option for this parameter is "USE_DEFAULTS" which only checks out all rows for feature classes. Check the help for the CreateReplica tool and see for yourself.
I hope this helps you get a better idea of what I am trying to do.
Any suggestions will be greatly appreciated.