What is the big deal about Sort? Sort is fundamental to the MapReduce
framework, the data is sorted between the Map and Reduce phases (see below).
Syncsort’s contribution allows native Hadoop sort to be replaced by an
alternative sort implementation, for both Map and Reduce sides, i.e. it
makes Sort phase pluggable.
Tendu Yogurtcu describes a couple of new use cases that the pluggable sort implementation contributed by Syncsort to Apache Hadoop is opening:
- Optimized sort implementations and full joins
- Hash-based aggregations with no sort requirements
- Reducers that can start before all Mappers complete
Original title and link: Use Cases for Hadoop’s New Pluggable Sort ( ©myNoSQL)