I designing clustered column store indexed fact tables.
Later on many new columns might be added to these currently ~90 columns tables.
How adding new nullable column is affected by the size (number of rows) of the table? Is it a metadata only operation, or some data reorganization goes on under the hood which i need to take into account?