The following happens:
- generate a bunch of random numbers
- write these to storage
- read them back from storage
- identify primes
- report the time everything took
Submission example:
spark-submit --master yarn --deploy-mode cluster spark-hadoop-performance-assembly-1.0.jar --records 1000000000 --partitions 792 --output /benchmark