Problem: 1.2 GB (14 Millions records) stored in an apache spark dataframe. The computation took less than 1 min but the writing to an MSSQL SERVER tabl