May 2018
Intermediate to advanced
256 pages
7h 6m
English
MapReduce is a model and opinionized implementation for analyzing and summarizing business and enterprise data using a distributed and massively parallel algorithm on a cluster of servers (such as HDFS clusters.) The map portion provides a programmatic way to filter and sort data, whereas the reduce performs the final summary operations. The model that ties MapReduce together also orchestrates the processing across the distributed servers. As HDFS clusters, this provides redundancy and fault tolerance to each task submitted and prevents data loss or extra delays based on execution stalls.
Read now
Unlock full access