September 2018
Intermediate to advanced
398 pages
9h 43m
English
The methods available in Dataset and the RDD are of two kinds:
When a job is triggered by an action method, it is divided into several stages. A stage is a part of a job that can be run without having to shuffle the data across different nodes of the cluster. It can encompass several transformations, such as map and filter. But as soon as one transformation, such as join, requires the data be moved (shuffling), another stage must be introduced.
The data contained in ...
Read now
Unlock full access