Warning: Eoulsan 2.x do not currently work on AWS. This page is outdated.

Configure Amazon Elastic MapReduce

Currently Eoulsan can only work on cloud with Amazon Web Service. Eoulsan use Amazon Elastic MapReduce service to execute an analysis on Cloud. This service provide on demand Hadoop clusters to customers. You can get more information about this service on Amazon MapReduce website.

Once you get an Amazon Web Service Account and sign-in for Amazon MapReduce and Amazon S3, go to Security credentials page of your AWS Account. In this page create or get your access key. You need a couple of valid access key/secret key to launch Eoulsan on Amazon Cloud. See the Amazon Elastic MapReduce Getting Started Guide for more information. Eoulsan use the AWS Java SDK library to transfer data to Amazon S3 and launch Amazon Elastic MapReduce jobs.

Eoulsan has been successfully tested with m1.xlarge instances and Hadoop 0.20, 0.20.205 and 1.03.

Configure Eoulsan to work with Amazon Elastic MapReduce

Eoulsan behaviour with Amazon Elastic MapReduce is the same as with a Hadoop cluster. You don't have to change your design file or the steps of the workflow file.

The only requierement is to set the AWS credentials in configuration file or workflow file. We strongly recommand to store this values in a configuration file. See the configuration file section for more information.

Launch Eoulsan in AWS mode

The launch of an Eoulsan analysis on AWS is described in AWS Elastic MapReduce exec command section.