config.file

#Config file is a tab-delimited file with these expected fields -
# field1 - EMR AMI version (may be a, a.b or a.b.c format or 'default', default will be used if exact match cannot be found)
# field2 - Spark build version (the REQUESTED_VERSION is matched against this field, if not specified it will look for the term 'default', default will be used if exact match cannot be found)
# field3 - (optional) shell program to use to execute the install script, defaults to python
# field4 - s3 location of a specific install script for this AMI and Spark version (is ran with environment variables S3SparkInstallPath, SparkBuild and Ec2Region)
# field5 - s3 path to files that will be used for Spark binaries, how this field is used is up to the install script, sets enviroment variable 'SparkS3InstallPath'
# field6 - (optional) s3 path to the script executed when "-x" argument is used, defaults to AWS provided version if empty
# field7 - (optional) s3 path to the script executed when "-g" argument is used to install ganlia, defaults to AWS provided version if empty
# field8 - (optional) s3 location for Scala binaries to install, how this field is used is up to the install script, set enviroment variable 'ScalaS3Location'
# field9 - (optional) s3 base path for supporting reference files, for example, where to find configure-spark.bash, this file would be expected to be found in this path, set environment variable 'SparkS3SupportingFilesPath'
#
default	2.0	python	s3://support.elasticmapreduce/spark/install-spark-script.py	s3://your-bucket-name/spark-2.0.0-preview-bin-hadoop2.7.tgz	s3://support.elasticmapreduce/spark/maximize-spark-default-config	s3://support.elasticmapreduce/spark/install-ganglia-metrics	s3://support.elasticmapreduce/spark/scala-2.11-builds/spark-1.2.1.a.tgz	s3://support.elasticmapreduce/spark/
#