I have small csv file with a counties dictionary which I want to use in a large spark job. I load this file via –files option of spark-submit: