06-26-2021 10:52 AM
It looks like the the spark memory is not sufficient to cache all the data so it read always from source
