Reasons why spark is faster than hadoop

(1) Spark uses a lot of memory

(2) Core data structure: RDD (Resilient Distributed Data Set), on which many computing functions are provided,

(3) The function implemented by the next map or reduce in the original hadoop can be divided into multiple jobs under spark. If hadoop is regarded as a big stone in a jar, then spark's job is a gravel in the jar, which can be Pack more.

（）