Vargrant provisioning files to spin up a modest (4 node) Hadoop
cluster for experiments processing HTRC Extracted Feature JSON files
suitable for ingesting into Solr.

Top-level code Apache Spark, processing HDFS stored JSON files, hence
the need for an underlying Hadoop cluster.

Provisioning based on the following online resources, but updated to
use newer versions of Ubuntu, Java, and Hadoop.

  http://cscarioni.blogspot.co.nz/2012/09/setting-up-hadoop-virtual-cluster-with.html

  https://github.com/calo81/vagrant-hadoop-cluster