Changes to allow EMR control through fog

* Better method to turn options hash into a k/v list
* run_job_flow and add_instance_groups added as methods on EMR
* Ability to fire Hive job added

lib/fog/aws.rb

@@ -19,6 +19,7 @@ module Fog
     service(:sns,             'aws/sns')
     service(:sqs,             'aws/sqs')
     service(:storage,         'aws/storage')
+    service(:emr,             'aws/emr')
 
     def self.indexed_param(key, values)
       params = {}
@@ -26,10 +27,34 @@ module Fog
         key << '.%d'
       end
       [*values].each_with_index do |value, index|
-        params[format(key, index + 1)] = value
+        if value.respond_to?('keys')
+          k = format(key, index + 1)
+          value.each do | vkey, vvalue |
+            params["#{k}.#{vkey}"] = vvalue
+          end
+        else
+          params[format(key, index + 1)] = value
+        end
       end
       params
     end
+    
+    def self.serialize_keys(key, value, options = {})
+      case value
+      when Hash
+        value.each do | k, v |
+          options.merge!(serialize_keys("#{key}.#{k}", v))
+        end
+        return options
+      when Array
+        value.each_with_index do | it, idx |
+          options.merge!(serialize_keys("#{key}.member.#{(idx + 1)}", it))
+        end
+        return options
+      else
+        return {key => value}
+      end
+    end
 
     def self.indexed_filters(filters)
       params = {}

lib/fog/aws/emr.rb

@@ -0,0 +1,133 @@
+require File.expand_path(File.join(File.dirname(__FILE__), '..', 'aws'))
+
+module Fog
+  module AWS
+    class EMR < Fog::Service
+
+      class IdentifierTaken < Fog::Errors::Error; end
+
+      requires :aws_access_key_id, :aws_secret_access_key
+      recognizes :region, :host, :path, :port, :scheme, :persistent
+
+      request_path 'fog/aws/requests/emr'
+
+      request :add_instance_groups
+#      request :add_job_flow_steps
+#      request :describe_job_flows
+#      request :modify_instance_groups
+      request :run_job_flow
+#      request :set_termination_protection
+#      request :terminate_job_flows
+
+      # model_path 'fog/aws/models/rds'
+      # model       :server
+      # collection  :servers
+      # model       :snapshot
+      # collection  :snapshots
+      # model       :parameter_group
+      # collection  :parameter_groups
+      # 
+      # model       :parameter
+      # collection  :parameters
+      # 
+      # model       :security_group
+      # collection  :security_groups
+
+      class Mock
+
+        def initialize(options={})
+          Fog::Mock.not_implemented
+        end
+
+      end
+
+      class Real
+
+        # Initialize connection to EMR
+        #
+        # ==== Notes
+        # options parameter must include values for :aws_access_key_id and
+        # :aws_secret_access_key in order to create a connection
+        #
+        # ==== Examples
+        #   emr = EMR.new(
+        #    :aws_access_key_id => your_aws_access_key_id,
+        #    :aws_secret_access_key => your_aws_secret_access_key
+        #   )
+        #
+        # ==== Parameters
+        # * options<~Hash> - config arguments for connection.  Defaults to {}.
+        #   * region<~String> - optional region to use, in ['eu-west-1', 'us-east-1', 'us-west-1'i, 'ap-southeast-1']
+        #
+        # ==== Returns
+        # * EMR object with connection to AWS.
+        def initialize(options={})
+          @aws_access_key_id      = options[:aws_access_key_id]
+          @aws_secret_access_key  = options[:aws_secret_access_key]
+          @connection_options     = options[:connection_options] || {}
+          @hmac = Fog::HMAC.new('sha256', @aws_secret_access_key)
+
+          options[:region] ||= 'us-east-1'
+          @host = options[:host] || 'elasticmapreduce.amazonaws.com'
+          @path       = options[:path]        || '/'
+          @persistent = options[:persistent]  || false
+          @port       = options[:port]        || 443
+          @scheme     = options[:scheme]      || 'https'
+          @connection = Fog::Connection.new("#{@scheme}://#{@host}:#{@port}#{@path}", @persistent, @connection_options)
+        end
+
+        def reload
+          @connection.reset
+        end
+
+        private
+
+        def request(params)
+          idempotent  = params.delete(:idempotent)
+          parser      = params.delete(:parser)
+
+          body = Fog::AWS.signed_params(
+            params,
+            {
+              :aws_access_key_id  => @aws_access_key_id,
+              :hmac               => @hmac,
+              :host               => @host,
+              :path               => @path,
+              :port               => @port,
+              :version            => '2009-03-31' #'2010-07-28'
+            }
+          )
+
+          begin
+            response = @connection.request({
+              :body       => body,
+              :expects    => 200,
+              :headers    => { 'Content-Type' => 'application/x-www-form-urlencoded' },
+              :idempotent => idempotent,
+              :host       => @host,
+              :method     => 'POST',
+              :parser     => parser
+            })
+          rescue Excon::Errors::HTTPStatusError => error
+            if match = error.message.match(/<Code>(.*)<\/Code>[\s\\\w]+<Message>(.*)<\/Message>/m)
+              # case match[1].split('.').last
+              # when 'DBInstanceNotFound', 'DBParameterGroupNotFound', 'DBSnapshotNotFound', 'DBSecurityGroupNotFound'
+              #   raise Fog::AWS::RDS::NotFound.slurp(error, match[2])
+              # when 'DBParameterGroupAlreadyExists'
+              #   raise Fog::AWS::RDS::IdentifierTaken.slurp(error, match[2])
+              # else
+              #   raise
+              # end
+              raise
+            else
+              raise
+            end
+          end
+
+          response
+        end
+
+      end
+    end
+  end
+end

lib/fog/aws/parsers/emr/add_instance_groups.rb

@@ -0,0 +1,32 @@
+module Fog
+  module Parsers
+    module AWS
+      module EMR
+
+        class AddInstanceGroups < Fog::Parsers::Base
+
+          def start_element(name, attrs = [])
+            super
+            case name
+            when 'InstanceGroupIds'
+              @in_instance_group_ids = true
+              @instance_groups_ids = []
+            end
+          end
+
+          def end_element(name)
+            case name
+            when 'JobFlowId'
+              @response[name] = value
+            when 'InstanceGroupIds'
+              @in_instance_group_ids = false
+              @response['InstanceGroupIds'] = @instance_group_ids
+            when 'InstanceGroupId'
+              @instance_group_ids << value
+            end
+          end
+        end
+      end
+    end
+  end
+end

lib/fog/aws/parsers/emr/run_job_flow.rb

@@ -0,0 +1,22 @@
+module Fog
+  module Parsers
+    module AWS
+      module EMR
+
+        class RunJobFlow < Fog::Parsers::Base
+
+          def start_element(name, attrs = [])
+            super
+          end
+
+          def end_element(name)
+            case name
+            when 'JobFlowId'
+              @response[name] = value
+            end
+          end
+        end
+      end
+    end
+  end
+end

lib/fog/aws/requests/emr/add_instance_groups.rb

@@ -0,0 +1,47 @@
+module Fog
+  module AWS
+    class EMR
+      class Real
+
+        require 'fog/aws/parsers/emr/add_instance_groups'
+
+        # adds an instance group to a running cluster
+        # http://docs.amazonwebservices.com/ElasticMapReduce/latest/API/API_AddInstanceGroups.html
+        # ==== Parameters
+        # * JobFlowId <~String> - Job flow in which to add the instance groups
+        # * InstanceGroups<~Array> - Instance Groups to add
+        #   * 'BidPrice'<~String> - Bid price for each Amazon EC2 instance in the instance group when launching nodes as Spot Instances, expressed in USD.
+        #   * 'InstanceCount'<~Integer> - Target number of instances for the instance group
+        #   * 'InstanceRole'<~String> - MASTER | CORE | TASK The role of the instance group in the cluster
+        #   * 'InstanceType'<~String> - The Amazon EC2 instance type for all instances in the instance group
+        #   * 'MarketType'<~String> - ON_DEMAND | SPOT Market type of the Amazon EC2 instances used to create a cluster node
+        #   * 'Name'<~String> - Friendly name given to the instance group.
+        # 
+        # ==== Returns
+        # * response<~Excon::Response>:
+        #   * body<~Hash>:
+        def add_instance_groups(job_flow_id, options={})
+          
+          if instance_groups = options.delete('InstanceGroups')
+            options.merge!(Fog::AWS.indexed_param('InstanceGroups.member.%d', [*instance_groups]))
+          end
+    
+          request({
+            'Action'  => 'CreateDBInstance',
+            'JobFlowId' => db_name,
+            :parser   => Fog::Parsers::AWS::EMR::AddInstanceGroups.new,
+          }.merge(options))
+        end
+
+      end
+
+      class Mock
+
+        def add_instance_groups(db_name, options={})
+          Fog::Mock.not_implemented
+        end
+
+      end
+    end
+  end
+end

lib/fog/aws/requests/emr/run_job_flow.rb

@@ -0,0 +1,106 @@
+module Fog
+  module AWS
+    class EMR
+      class Real
+
+        require 'fog/aws/parsers/emr/run_job_flow'
+
+        # creates and starts running a new job flow
+        # http://docs.amazonwebservices.com/ElasticMapReduce/latest/API/API_RunJobFlow.html
+        # ==== Parameters
+        # * AdditionalInfo <~String> - A JSON string for selecting additional features.
+        # * BootstrapActions <~Array> - A list of bootstrap actions that will be run before Hadoop is started on the cluster nodes.
+        #   * 'Name'<~String> - The name of the bootstrap action
+        #   * 'ScriptBootstrapAction'<~Array> - The script run by the bootstrap action
+        #     * 'Args' <~Array> - A list of command line arguments to pass to the bootstrap action script
+        #     * 'Path' <~String> - Location of the script to run during a bootstrap action. Can be either a location in Amazon S3 or on a local file system.
+        # * Instances <~Array> - A specification of the number and type of Amazon EC2 instances on which to run the job flow.
+        #   * 'Ec2KeyName'<~String> - Specifies the name of the Amazon EC2 key pair that can be used to ssh to the master node as the user called "hadoop.
+        #   * 'HadoopVersion'<~String> - "0.18" | "0.20" Specifies the Hadoop version for the job flow
+        #   * 'InstanceCount'<~Integer> - The number of Amazon EC2 instances used to execute the job flow
+        #   * 'InstanceGroups'<~Array> - Configuration for the job flow's instance groups
+        #     * 'BidPrice' <~String> - Bid price for each Amazon EC2 instance in the instance group when launching nodes as Spot Instances, expressed in USD.
+        #     * 'InstanceCount'<~Integer> - Target number of instances for the instance group
+        #     * 'InstanceRole'<~String> - MASTER | CORE | TASK The role of the instance group in the cluster
+        #     * 'InstanceType'<~String> - The Amazon EC2 instance type for all instances in the instance group
+        #     * 'MarketType'<~String> - ON_DEMAND | SPOT Market type of the Amazon EC2 instances used to create a cluster node
+        #     * 'Name'<~String> - Friendly name given to the instance group.
+        #   * 'KeepJobFlowAliveWhenNoSteps' <~Boolean> - Specifies whether the job flow should terminate after completing all steps
+        #   * 'MasterInstanceType'<~String> - The EC2 instance type of the master node
+        #   * 'Placement'<~Array> - Specifies the Availability Zone the job flow will run in
+        #     * 'AvailabilityZone' <~String> - The Amazon EC2 Availability Zone for the job flow.
+        #   * 'SlaveInstanceType'<~String> - The EC2 instance type of the slave nodes
+        #   * 'TerminationProtected'<~Boolean> - Specifies whether to lock the job flow to prevent the Amazon EC2 instances from being terminated by API call, user intervention, or in the event of a job flow error
+        # * LogUri <~String> - Specifies the location in Amazon S3 to write the log files of the job flow. If a value is not provided, logs are not created
+        # * Name <~String> - The name of the job flow
+        # * Steps <~Array> - A list of steps to be executed by the job flow
+        #   * 'ActionOnFailure'<~String> - TERMINATE_JOB_FLOW | CANCEL_AND_WAIT | CONTINUE Specifies the action to take if the job flow step fails
+        #   * 'HadoopJarStep'<~Array> - Specifies the JAR file used for the job flow step
+        #     * 'Args'<~String list> - A list of command line arguments passed to the JAR file's main function when executed.
+        #     * 'Jar'<~String> - A path to a JAR file run during the step.
+        #     * 'MainClass'<~String> - The name of the main class in the specified Java file. If not specified, the JAR file should specify a Main-Class in its manifest file
+        #     * 'Properties'<~Array> - A list of Java properties that are set when the step runs. You can use these properties to pass key value pairs to your main function
+        #       * 'Key'<~String> - The unique identifier of a key value pair
+        #       * 'Value'<~String> - The value part of the identified key
+        #   * 'Name'<~String> - The name of the job flow step
+        # 
+        # ==== Returns
+        # * response<~Excon::Response>:
+        #   * body<~Hash>:
+        def run_job_flow(name, options={})
+          
+          if bootstrap_actions = options.delete('BootstrapActions')
+            options.merge!(Fog::AWS.serialize_keys('BootstrapActions', bootstrap_actions))
+          end
+          
+          if instances = options.delete('Instances')
+            options.merge!(Fog::AWS.serialize_keys('Instances', instances))
+          end
+          
+          if steps = options.delete('Steps')
+            options.merge!(Fog::AWS.serialize_keys('Steps', steps))
+          end
+    
+          request({
+            'Action'  => 'RunJobFlow',
+            'Name' => name,
+            :parser   => Fog::Parsers::AWS::EMR::RunJobFlow.new,
+          }.merge(options))
+        end
+        
+        def run_hive(name, options={})
+          steps = []
+          steps << {
+            'Name' => 'Setup Hive',
+            'HadoopJarStep' => {
+              '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']},
+            'ActionOnFailure' => 'TERMINATE_JOB_FLOW'
+          }
+          steps << {
+            'Name' => 'Install Hive Site Configuration',
+            'HadoopJarStep' => {
+              '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-site', '--hive-site=s3://raybeam.okl/prod/hive/hive-site.xml']},
+            'ActionOnFailure' => 'TERMINATE_JOB_FLOW'
+          }
+          options['Steps'] = steps
+          
+          if not options['Instances'].nil?
+            options['Instances']['KeepJobFlowAliveWhenNoSteps'] = true
+          end
+          
+          run_job_flow name, options
+        end
+      end
+
+      class Mock
+
+        def run_job_flow(db_name, options={})
+          Fog::Mock.not_implemented
+        end
+
+      end
+    end
+  end
+end