Module: Wukong::Elasticsearch

Includes:
Plugin
Defined in:
lib/wonderdog.rb,
lib/wonderdog/timestamp.rb,
lib/wonderdog/index_and_mapping.rb,
lib/wonderdog/hadoop_invocation_override.rb

Overview

Wonderdog provides Java code that couples Hadoop streaming to Wukong. This module adds some overrides which enables the wu-hadoop program to leverage this code.

Defined Under Namespace

Modules: HadoopInvocationOverride Classes: IndexAndMapping, Timestamp

Class Method Summary collapse

Class Method Details

.boot(settings, root) ⇒ Object

Boot Wonderdog with the given settings in the given dir.

Parameters:

  • settings (Configliere::Param)
  • root (String) 


38
39
 
# File 'lib/wonderdog.rb', line 38

def self.boot settings, root
end

.configure(settings, program) ⇒ Configliere::Param

Configure the given settings to be able to work with Elasticsearch.

Parameters:

  • settings (Configliere::Param)

Returns:

  • (Configliere::Param) —

    the newly configured settings



16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
 
# File 'lib/wonderdog.rb', line 16

def self.configure settings, program
  return unless program == 'wu-hadoop'
  settings.define(:es_tmp_dir,        :description => "Temporary directory on the HDFS to store job files while reading/writing to ElasticSearch", :default => "/user/#{ENV['USER']}/wukong", :wukong_hadoop => true)
  settings.define(:es_lib_dir,        :description => "Directory containing Elasticsearch, Wonderdog, and other support jars", :default => "/usr/lib/hadoop/lib", :wukong_hadoop => true)
  settings.define(:es_config,         :description => "Where to find configuration files detailing how to join an ElasticSearch cluster", :wukong_hadoop => true)
  settings.define(:es_input_splits,   :description => "Number of input splits to target when reading from ElasticSearch", :type => Integer, :wukong_hadoop => true)
  settings.define(:es_request_size,   :description => "Number of objects requested during each batch read from ElasticSearch", :type => Integer, :wukong_hadoop => true)
  settings.define(:es_scroll_timeout, :description => "Amount of time to wait on a scroll", :wukong_hadoop => true)
  settings.define(:es_index_field,    :description => "Field to use from each record to override the default index", :wukong_hadoop => true)
  settings.define(:es_mapping_field,  :description => "Field to use from each record to override the default mapping", :wukong_hadoop => true)
  settings.define(:es_id_field,       :description => "If this field is present in a record, make an update request, otherwise make a create request", :wukong_hadoop => true)
  settings.define(:es_bulk_size,      :description => "Number of requests to batch locally before making a request to ElasticSearch", :type => Integer, :wukong_hadoop => true)
  settings.define(:es_query,          :description => "Query to use when defining input splits for ElasticSearch input",    :wukong_hadoop => true)
  settings.define(:es_transport,      :description => "Use a transport client to an existing node instead of spinning up a new node", :default => true, type: :boolean, :wukong_hadoop => true)
  settings.define(:es_transport_host, :description => "Host of existing node for transport client", default: 'localhost', :wukong_hadoop => true)
  settings.define(:es_transport_port, :description => "Port of existing node for transport client", default: 9300, type: Integer, :wukong_hadoop => true)
end