Q1
Single choice
You have an embarrassingly parallel or distributed batch job with a large amount of data running using Data Science Jobs.
What would be the best approach to run the workload?