Hello, Ask TOM Team.

I have a table A with 7 columns. I need to add two more columns (x, y) to this table or create a new one with a FK. These new columns could be "filled" with an insert or an update.

Business: We get XML invoices from customers. When XMLs arrives, they must be validated (schemas and business rules), meanwhile they are in *being processed state*. Then, when validation is finished (~1 sec), the application update the invoice state field in table A.

The insert method:
It would require create table B with three columns (PFK, x, y) because at the moment of insertion in table A I do not have the data to insert. So the app would insert in this child table.

The update method:
Add columns (x, y) to table A and then update them with an existing state update (mentioned above) in the application (I mean two more columns to the SET operator).

What is recommended for best practice: insert or update?

Thanks,