[bsg75]

[Docker] "containers over Spark" for scaling data science.

These are very different approaches. How would this comparison have been made?