Dynamo DB backups using the Ruby AWS SDK

dynamo-backups.rb

require 'aws-sdk'

AWS.config({:access_key_id => "your-access-key-id",
            :secret_access_key => "your-secrect-access-key",
            :region => "eu-west-1"}) #region is optional

log_uri = "s3n://dynamo-backups/dynamo-backup-resources/logs" #Can be anywhere you like on S3

emr = AWS::EMR.new
job = emr.job_flows.create('dynamo-db-backup', {log_uri: log_uri,
                                                       instances: 
                                                          {keep_job_flow_alive_when_no_steps: false,
                                                           instance_count: 1, 
                                                           master_instance_type: 'm1.small', 
                                                           slave_instance_type: 'm1.small'},
                                                        ami_version: "latest"})

job.add_steps([{:name => 'setup-hive',
                :action_on_failure => 'TERMINATE_JOB_FLOW',
                :hadoop_jar_step => {
                   :jar => "s3://us-east-1.elasticmapreduce/libs/script-runner/script-runner.jar",
                   :args => ["s3://us-east-1.elasticmapreduce/libs/hive/hive-script",
                             "--base-path",
                             "s3://us-east-1.elasticmapreduce/libs/hive/",
                             "--install-hive",
                             "--hive-versions",
                             "0.8.1.7"]}},  
               {:name => 'run-hive-script',
                :action_on_failure => 'CANCEL_AND_WAIT',
                :hadoop_jar_step => {
                  :jar => "s3://us-east-1.elasticmapreduce/libs/script-runner/script-runner.jar",
                  :args => ["s3://us-east-1.elasticmapreduce/libs/hive/hive-script",
                            "--base-path",
                            "s3://us-east-1.elasticmapreduce/libs/hive/",
                            "--hive-versions",
                            "0.8.1.7",
                            "--run-hive-script",
                            "--args",
                            "-f",
                            "s3n://dynamo-backup-resources/backup-to-s3.hql", 
                            "-hiveconf",
                            "backup_path=s3n://dynamo-backups/#{Time.now.strftime("%Y-%m-%d")}"
                            ]}}])
                            
STDOUT.puts "Job #{job.id} has been started on the EMR cluster. Please check the Amazon dashboard for progress. Logs are available at #{log_uri}/#{job.id}"