It's not clear to me how to set autocommit off/false. Is this a setting I'm supposed to make in my Spark configurations when I launch it in either -shell/-submit? Or is there some standard JDBC config file somewhere that gets referenced for these kinds of settings being turned off? I ask because the only documentation I see related to turning autocommit off is directly with the java.sql.Connection class, but I am trying to write using df.write.format("jdbc").options(Map( ... )).save() and it throws an error if you attempt to set an autocommit option. I also do not see an autocommit-related option in the Simba JBDC documentation where it has the list of available driver options that can be set via the options(Map( ... )) call. I can't use a Connection object as that only passes direct SQL commands and won't write out a DataFrame; I could loop it and inject the inserts, but that will be too clunky and slow.