I have a target table within an ETL mapping.
Pipe A, performs an update to fieldA. Pipe B I need to delete rows based on the value in fieldA (as well as a few other fields)
To perform the update, fieldA cannot be a PK. To perform the delete, fieldA is required to be a PK.
I'm trying to stay away from post-mapping SQL if I can so am seeking other options that may be available to me.
To perform update or delete, the actual physical table does not need to have Primary key, only the definition in Informatica should have the keys. I believe you can create two separate target definitions of the same table, and define different keys. You can then use those separate target definitions in the two pipelines.