Inside a data flow, is it possible to do a lookup, insert the non-matching rows, re-lookup and then continue with the full (original) data set?
I can't see how you get/re-use the inserted rows after they've gone to an OLE DB Destination
, nor can I see that a Multicast
could wait until a lookup insert branch completes (and then union all them back-in).
If it can't be done inside the data flow, would I have to repeat all my filtering logic in a previous data flow task simply to do the lookup inserts first?
This might be an obvious/already answered question, but I can't find much after searching.