The integration of the
SAS Embedded Process and Hadoop allows scoring code to be run directly
on Hadoop using the SAS Embedded Process.
The SAS Embedded Process
is a SAS server process that runs inside Hadoop to read and write
data. A model publishing macro creates scoring files and stores them
in a Hadoop Distributed File System (HDFS) directory. These scoring
files are then used by a Hadoop MapReduce function to run the scoring
model.
You can create traditional
scoring models by using the SAS Enterprise Miner Score Code Export
node. In the July 2015 release of SAS 9.4, the SAS Scoring Accelerator
for Hadoop supports analytic store scoring. SAS Factory Miner components,
the HPFOREST and HPSVM components, generate the analytic store file
and the SAS scoring model program, and format catalog that are used
in analytic store scoring.
The SAS Scoring Accelerator
for Hadoop requires a specific version of Hadoop. For more information,
see the SAS Foundation system requirements documentation for your
operating environment.