In migration effort we're being asked to optimize our code because our new data source has billions of rows versus our million row subsets which is being taken away.
To optimize we need to do what was being done for us previously and that is retrieve our subset of data. To that end we've been asked to only retrieve columns from a given table that we actually use.
The only way I can envision that is to create a stored procedure to query the metadata of a given table and then do a search for each column. I was hoping to find someone who actually had such a utility they would be willing to share as I'm under a severe time constraint.
