I have an input csv which feeds eventually into 10 different tables. Is it better (efficient) to have 10 output ports from the DelimitedReader or to have a SimpleCopy node to split the records for insertion into DB_OUTPUT_TABLE nodes?
Along these lines, if I have to set phases on the DB_OUTPUT_TABLE nodes to manage some FK constraints, does the matter in which I read/copy the input data matter across Phases?
Thanks,
Brad