Unfortunately, as I didn't receive feedback from the community to give me guidance, I had to rack my brains a lot, hours and hours of testing, but I managed to do what I wanted. I downloaded Spark in the same version as cdh 6.3.4, I configured the spark configuration files with the information from cdh 6.3.4, so when calling "spark-submit" the job is executed in the cdh cluster
... View more
