This is an known problem with spark and databrick jdbc jar, and databrick has a workaround: https://kb.databricks.com/python/column-value-errors-when-connecting-from-apache-spark-to-databricks-using-spark-jdbc