I have 2 nodes (datetime to string, and string to datetime) I want the 2nd node only to act on the columns acted on in the first. This is running inside a variable loop so that I can act on several tables.
how do you control which columns are converted in the first node? And also what are you really doing here? From Date&Time to String and back? Maybe there is a different way to deal with it?
Described behavior can be due to Enforce exclusion/inclusion configuration in second node. If you have Enforce exclusion option and empty column list you’ll always get all columns converted regardless of flow variable. I found this topic where I (maybe) explained it a bit more with couple of other ideas/remarks: