Class: Fluent::Plugin::ElasticsearchOutput

Inherits:
Output
  • Object
show all
Includes:
ElasticsearchIndexTemplate, ElasticsearchConstants, ElasticsearchIndexLifecycleManagement, ElasticsearchTLS
Defined in:
lib/fluent/plugin/out_elasticsearch.rb

Defined Under Namespace

Classes: MissingIdFieldError, RecoverableRequestFailure, RequestInfo, RetryStreamEmitFailure, RetryStreamError, UnrecoverableRequestFailure

Constant Summary collapse

DEFAULT_BUFFER_TYPE = 
"memory"
DEFAULT_ELASTICSEARCH_VERSION =

For compatibility.

5
DEFAULT_TYPE_NAME_ES_7x = 
"_doc".freeze
DEFAULT_TYPE_NAME = 
"fluentd".freeze
DEFAULT_RELOAD_AFTER = 
-1
DEFAULT_TARGET_BULK_BYTES = 
-1
DEFAULT_POLICY_ID = 
"logstash-policy"

Constants included from ElasticsearchTLS

Fluent::Plugin::ElasticsearchTLS::DEFAULT_VERSION, Fluent::Plugin::ElasticsearchTLS::SUPPORTED_TLS_VERSIONS

Constants included from ElasticsearchIndexLifecycleManagement

Fluent::Plugin::ElasticsearchIndexLifecycleManagement::ILM_DEFAULT_POLICY_PATH

Constants included from ElasticsearchConstants

Fluent::Plugin::ElasticsearchConstants::BODY_DELIMITER, Fluent::Plugin::ElasticsearchConstants::CREATE_OP, Fluent::Plugin::ElasticsearchConstants::ID_FIELD, Fluent::Plugin::ElasticsearchConstants::INDEX_OP, Fluent::Plugin::ElasticsearchConstants::TIMESTAMP_FIELD, Fluent::Plugin::ElasticsearchConstants::UPDATE_OP, Fluent::Plugin::ElasticsearchConstants::UPSERT_OP

Instance Attribute Summary collapse

Instance Method Summary collapse

Methods included from ElasticsearchTLS

included, #set_tls_minmax_version_config

Methods included from ElasticsearchIndexLifecycleManagement

#create_ilm_policy, #default_policy_payload, #get_ilm_policy, #ilm_policy_exists?, #ilm_policy_put, #setup_ilm, #verify_ilm_working, #xpack_info

Methods included from ElasticsearchIndexTemplate

#create_rollover_alias, #get_custom_template, #get_template, #get_template_name, #host_unreachable_exceptions, #indexcreation, #inject_ilm_settings_to_template, #retry_operate, #rollover_alias_payload, #template_custom_install, #template_exists?, #template_install, #template_put, #templates_hash_install

Constructor Details

#initialize ⇒ ElasticsearchOutput

Returns a new instance of ElasticsearchOutput.



196
197
198
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 196

def initialize
  super
end

Instance Attribute Details

#alias_indexes ⇒ Object (readonly)

Returns the value of attribute alias_indexes.



59
60
61
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 59

def alias_indexes
  @alias_indexes
end

#api_key_header ⇒ Object (readonly)

Returns the value of attribute api_key_header.



63
64
65
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 63

def api_key_header
  @api_key_header
end

#compressable_connection ⇒ Object (readonly)

Returns the value of attribute compressable_connection.



62
63
64
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 62

def compressable_connection
  @compressable_connection
end

#ssl_version_options ⇒ Object (readonly)

Returns the value of attribute ssl_version_options.



61
62
63
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 61

def ssl_version_options
  @ssl_version_options
end

#template_names ⇒ Object (readonly)

Returns the value of attribute template_names.



60
61
62
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 60

def template_names
  @template_names
end

Instance Method Details

#append_record_to_messages(op, meta, header, record, msgs) ⇒ Object

append_record_to_messages adds a record to the bulk message payload to be submitted to Elasticsearch. Records that do not include ‘_id’ field are skipped when ‘write_operation’ is configured for ‘create’ or ‘update’

returns ‘true’ if record was appended to the bulk message

and 'false' otherwise



730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 730

def append_record_to_messages(op, meta, header, record, msgs)
  case op
  when UPDATE_OP, UPSERT_OP
    if meta.has_key?(ID_FIELD)
      header[UPDATE_OP] = meta
      msgs << @dump_proc.call(header) << BODY_DELIMITER
      msgs << @dump_proc.call(update_body(record, op)) << BODY_DELIMITER
      return true
    end
  when CREATE_OP
    if meta.has_key?(ID_FIELD)
      header[CREATE_OP] = meta
      msgs << @dump_proc.call(header) << BODY_DELIMITER
      msgs << @dump_proc.call(record) << BODY_DELIMITER
      return true
    end
  when INDEX_OP
    header[INDEX_OP] = meta
    msgs << @dump_proc.call(header) << BODY_DELIMITER
    msgs << @dump_proc.call(record) << BODY_DELIMITER
    return true
  end
  return false
end

#backend_options ⇒ Object 



471
472
473
474
475
476
477
478
479
480
481
482
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 471

def backend_options
  case @http_backend
  when :excon
    { client_key: @client_key, client_cert: @client_cert, client_key_pass: @client_key_pass, nonblock: @http_backend_excon_nonblock }
  when :typhoeus
    require 'faraday/typhoeus'
    { sslkey: @client_key, sslcert: @client_cert, keypasswd: @client_key_pass }
  end
rescue LoadError => ex
  log.error_backtrace(ex.backtrace)
  raise Fluent::ConfigError, "You must install #{@http_backend} gem. Exception: #{ex}"
end

#client(host = nil, compress_connection = false) ⇒ Object 



590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 590

def client(host = nil, compress_connection = false)
  return cloud_client if @cloud_id

  # check here to see if we already have a client connection for the given host
  connection_options = get_connection_options(host)

  @_es = nil unless is_existing_connection(connection_options[:hosts])
  @_es = nil unless @compressable_connection == compress_connection

  @_es ||= begin
    @compressable_connection = compress_connection
    @current_config = connection_options[:hosts].clone
    adapter_conf = lambda {|f| f.adapter @http_backend, @backend_options }
    local_reload_connections = @reload_connections
    if local_reload_connections && @reload_after > DEFAULT_RELOAD_AFTER
      local_reload_connections = @reload_after
    end

    gzip_headers = if compress_connection
                     {'Content-Encoding' => 'gzip'}
                   else
                     {}
                   end
    headers = { 'Content-Type' => @content_type.to_s }
                .merge(@custom_headers)
                .merge(@api_key_header)
                .merge(gzip_headers)
    ssl_options = { verify: @ssl_verify, ca_file: @ca_file}.merge(@ssl_version_options)

    transport = TRANSPORT_CLASS::Transport::HTTP::Faraday.new(connection_options.merge(
                                                                        options: {
                                                                          reload_connections: local_reload_connections,
                                                                          reload_on_failure: @reload_on_failure,
                                                                          resurrect_after: @resurrect_after,
                                                                          logger: @transport_logger,
                                                                          transport_options: {
                                                                            headers: headers,
                                                                            request: { timeout: @request_timeout },
                                                                            ssl: ssl_options,
                                                                          },
                                                                          http: {
                                                                            user: @user,
                                                                            password: @password,
                                                                            scheme: @scheme
                                                                          },
                                                                          sniffer_class: @sniffer_class,
                                                                          serializer_class: @serializer_class,
                                                                          selector_class: @selector_class,
                                                                          compression: compress_connection,
                                                                        }), &adapter_conf)
    Elasticsearch::Client.new transport: transport
  end
end

#client_library_version ⇒ Object 



513
514
515
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 513

def client_library_version
  Elasticsearch::VERSION
end

#cloud_client ⇒ Object 



582
583
584
585
586
587
588
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 582

def cloud_client
  Elasticsearch::Client.new(
    cloud_id: @cloud_id,
    user: @user,
    password: @password
  )
end

#compression ⇒ Object 



454
455
456
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 454

def compression
  !(@compression_level == :no_compression)
end

#compression_strategy ⇒ Object 



458
459
460
461
462
463
464
465
466
467
468
469
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 458

def compression_strategy
  case @compression_level
  when :default_compression
    Zlib::DEFAULT_COMPRESSION
  when :best_compression
    Zlib::BEST_COMPRESSION
  when :best_speed
    Zlib::BEST_SPEED
  else
    Zlib::NO_COMPRESSION
  end
end

#configure(conf) ⇒ Object

Raises:

  • (Fluent::ConfigError) 


200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 200

def configure(conf)
  compat_parameters_convert(conf, :buffer)

  super
  if placeholder_substitution_needed_for_template?
    # nop.
  elsif not @buffer_config.chunk_keys.include? "tag" and
    not @buffer_config.chunk_keys.include? "_index"
    raise Fluent::ConfigError, "'tag' or '_index' in chunk_keys is required."
  end
  @time_parser = create_time_parser
  @backend_options = backend_options
  @ssl_version_options = set_tls_minmax_version_config(@ssl_version, @ssl_max_version, @ssl_min_version)

  if @remove_keys
    @remove_keys = @remove_keys.split(/\s*,\s*/)
  end

  if @target_index_key && @target_index_key.is_a?(String)
    @target_index_key = @target_index_key.split '.'
  end

  if @target_type_key && @target_type_key.is_a?(String)
    @target_type_key = @target_type_key.split '.'
  end

  if @remove_keys_on_update && @remove_keys_on_update.is_a?(String)
    @remove_keys_on_update = @remove_keys_on_update.split ','
  end

  @api_key_header = setup_api_key

  raise Fluent::ConfigError, "'max_retry_putting_template' must be greater than or equal to zero." if @max_retry_putting_template < 0
  raise Fluent::ConfigError, "'max_retry_get_es_version' must be greater than or equal to zero." if @max_retry_get_es_version < 0

  # Dump log when using host placeholders and template features at same time.
  valid_host_placeholder = placeholder?(:host_placeholder, @host)
  if valid_host_placeholder && (@template_name && @template_file || @templates)
    if @verify_es_version_at_startup
      raise Fluent::ConfigError, "host placeholder, template installation, and verify Elasticsearch version at startup are exclusive feature at same time. Please specify verify_es_version_at_startup as `false` when host placeholder and template installation are enabled."
    end
    log.info "host placeholder and template installation makes your Elasticsearch cluster a bit slow down(beta)."
  end

  raise Fluent::ConfigError, "You can't specify ilm_policy and ilm_policies at the same time" unless @ilm_policy.empty? or @ilm_policies.empty?

  unless @ilm_policy.empty?
    @ilm_policies = { @ilm_policy_id => @ilm_policy }
  end
  @alias_indexes = []
  @template_names = []
  if !dry_run?
    if @template_name && @template_file
      if @enable_ilm
        raise Fluent::ConfigError, "deflector_alias is prohibited to use with enable_ilm at same time." if @deflector_alias
      end
      if @ilm_policy.empty? && @ilm_policy_overwrite
        raise Fluent::ConfigError, "ilm_policy_overwrite requires a non empty ilm_policy."
      end
      if @logstash_format || placeholder_substitution_needed_for_template?
        class << self
          alias_method :template_installation, :template_installation_actual
        end
      else
        template_installation_actual(@deflector_alias ? @deflector_alias : @index_name, @template_name, @customize_template, @application_name, @index_name, @ilm_policy_id)
      end
      verify_ilm_working if @enable_ilm
    end
    if @templates
      retry_operate(@max_retry_putting_template,
                    @fail_on_putting_template_retry_exceed,
                    @catch_transport_exception_on_retry) do
        templates_hash_install(@templates, @template_overwrite)
      end
    end
  end

  @truncate_mutex = Mutex.new
  if @truncate_caches_interval
    timer_execute(:out_elasticsearch_truncate_caches, @truncate_caches_interval) do
      log.info('Clean up the indices and template names cache')

      @truncate_mutex.synchronize {
        @alias_indexes.clear
        @template_names.clear
      }
    end
  end

  @serializer_class = nil
  begin
    require 'oj'
    @dump_proc = Oj.method(:dump)
    if @prefer_oj_serializer
      @serializer_class = Fluent::Plugin::Serializer::Oj
      Elasticsearch::API.settings[:serializer] = Fluent::Plugin::Serializer::Oj
    end
  rescue LoadError
    @dump_proc = Yajl.method(:dump)
  end

  raise Fluent::ConfigError, "`cloud_auth` must be present if `cloud_id` is present" if @cloud_id && @cloud_auth.nil?
  raise Fluent::ConfigError, "`password` must be present if `user` is present" if @user && @password.nil?

  if @cloud_auth
    @user = @cloud_auth.split(':', -1)[0]
    @password = @cloud_auth.split(':', -1)[1]
  end

  if @user && m = @user.match(/%{(?<user>.*)}/)
    @user = URI.encode_www_form_component(m["user"])
  end
  if @password && m = @password.match(/%{(?<password>.*)}/)
    @password = URI.encode_www_form_component(m["password"])
  end

  @transport_logger = nil
  if @with_transporter_log
    @transport_logger = log
    log_level = conf['@log_level'] || conf['log_level']
    log.warn "Consider to specify log_level with @log_level." unless log_level
  end
  # Specify @sniffer_class before calling #client.
  # #detect_es_major_version uses #client.
  @sniffer_class = nil
  begin
    @sniffer_class = Object.const_get(@sniffer_class_name) if @sniffer_class_name
  rescue Exception => ex
    raise Fluent::ConfigError, "Could not load sniffer class #{@sniffer_class_name}: #{ex}"
  end

  @selector_class = nil
  begin
    @selector_class = Object.const_get(@selector_class_name) if @selector_class_name
  rescue Exception => ex
    raise Fluent::ConfigError, "Could not load selector class #{@selector_class_name}: #{ex}"
  end

  @last_seen_major_version = if major_version = handle_last_seen_es_major_version
                               major_version
                             else
                               @default_elasticsearch_version
                             end
  if @suppress_type_name && @last_seen_major_version >= 7
    @type_name = nil
  else
    if @last_seen_major_version == 6 && @type_name != DEFAULT_TYPE_NAME_ES_7x
      log.info "Detected ES 6.x: ES 7.x will only accept `_doc` in type_name."
    end
    if @last_seen_major_version == 7 && @type_name != DEFAULT_TYPE_NAME_ES_7x
      log.warn "Detected ES 7.x: `_doc` will be used as the document `_type`."
      @type_name = '_doc'.freeze
    end
    if @last_seen_major_version >= 8 && @type_name != DEFAULT_TYPE_NAME_ES_7x
      log.debug "Detected ES 8.x or above: This parameter has no effect."
      @type_name = nil
    end
  end

  if @validate_client_version && !dry_run?
    if @last_seen_major_version != client_library_version.to_i
      raise Fluent::ConfigError, <<-EOC
        Detected ES #{@last_seen_major_version} but you use ES client #{client_library_version}.
        Please consider to use #{@last_seen_major_version}.x series ES client.
      EOC
    end
  end

  if @last_seen_major_version >= 6
    case @ssl_version
    when :SSLv23, :TLSv1, :TLSv1_1
      if @scheme == :https
        log.warn "Detected ES 6.x or above and enabled insecure security:
                  You might have to specify `ssl_version TLSv1_2` in configuration."
      end
    end
  end

  if @ssl_version && @scheme == :https
    if !@http_backend_excon_nonblock
      log.warn "TLS handshake will be stucked with block connection.
                Consider to set `http_backend_excon_nonblock` as true"
    end
  end

  # Consider missing the prefix of "$." in nested key specifiers.
  @id_key = convert_compat_id_key(@id_key) if @id_key
  @parent_key = convert_compat_id_key(@parent_key) if @parent_key
  @routing_key = convert_compat_id_key(@routing_key) if @routing_key

  @routing_key_name = configure_routing_key_name
  @meta_config_map = create_meta_config_map
  @current_config = nil
  @compressable_connection = false

  @ignore_exception_classes = @ignore_exceptions.map do |exception|
    unless Object.const_defined?(exception)
      log.warn "Cannot find class #{exception}. Will ignore it."

      nil
    else
      Object.const_get(exception)
    end
  end.compact

  if @bulk_message_request_threshold < 0
    class << self
      alias_method :split_request?, :split_request_size_uncheck?
    end
  else
    class << self
      alias_method :split_request?, :split_request_size_check?
    end
  end

  if Gem::Version.create(::TRANSPORT_CLASS::VERSION) < Gem::Version.create("7.2.0")
    if compression
      raise Fluent::ConfigError, <<-EOC
        Cannot use compression with elasticsearch-transport plugin version < 7.2.0
        Your elasticsearch-transport plugin version version is #{TRANSPORT_CLASS::VERSION}.
        Please consider to upgrade ES client.
      EOC
    end
  end
end

#configure_routing_key_name ⇒ Object 



517
518
519
520
521
522
523
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 517

def configure_routing_key_name
  if @last_seen_major_version >= 7
    'routing'
  else
    '_routing'
  end
end

#connection_options_description(con_host = nil) ⇒ Object 



715
716
717
718
719
720
721
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 715

def connection_options_description(con_host=nil)
  get_connection_options(con_host)[:hosts].map do |host_info|
    attributes = host_info.dup
    attributes[:password] = 'obfuscated' if attributes.has_key?(:password)
    attributes.inspect
  end.join(', ')
end

#convert_compat_id_key(key) ⇒ Object 



525
526
527
528
529
530
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 525

def convert_compat_id_key(key)
  if key.include?('.') && !key.start_with?('$[')
    key = "$.#{key}" unless key.start_with?('$.')
  end
  key
end

#convert_numeric_time_into_string(numeric_time, time_key_format = "%Y-%m-%d %H:%M:%S.%N%z") ⇒ Object 



570
571
572
573
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 570

def convert_numeric_time_into_string(numeric_time, time_key_format = "%Y-%m-%d %H:%M:%S.%N%z")
  numeric_time_parser = Fluent::NumericTimeParser.new(:float)
  Time.at(numeric_time_parser.parse(numeric_time).to_r).strftime(time_key_format)
end

#create_meta_config_map ⇒ Object 



532
533
534
535
536
537
538
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 532

def create_meta_config_map
  result = []
  result << [record_accessor_create(@id_key), '_id'] if @id_key
  result << [record_accessor_create(@parent_key), '_parent'] if @parent_key
  result << [record_accessor_create(@routing_key), @routing_key_name] if @routing_key
  result
end

#create_time_parser ⇒ Object

once fluent v0.14 is released we might be able to use Fluent::Parser::TimeParser, but it doesn’t quite do what we want - if gives

sec,nsec

where as we want something we can call ‘strftime` on…



543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 543

def create_time_parser
  if @time_key_format
    begin
      # Strptime doesn't support all formats, but for those it does it's
      # blazingly fast.
      strptime = Strptime.new(@time_key_format)
      Proc.new { |value|
        value = convert_numeric_time_into_string(value, @time_key_format) if value.is_a?(Numeric)
        strptime.exec(value).to_datetime
      }
    rescue
      # Can happen if Strptime doesn't recognize the format; or
      # if strptime couldn't be required (because it's not installed -- it's
      # ruby 2 only)
      Proc.new { |value|
        value = convert_numeric_time_into_string(value, @time_key_format) if value.is_a?(Numeric)
        DateTime.strptime(value, @time_key_format)
      }
    end
  else
    Proc.new { |value|
      value = convert_numeric_time_into_string(value) if value.is_a?(Numeric)
      DateTime.parse(value)
    }
  end
end

#detect_es_major_version ⇒ Object 



496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 496

def detect_es_major_version
  begin
    @_es_info ||= client.info
  rescue ::Elasticsearch::UnsupportedProductError => e
    raise Fluent::ConfigError, "Using Elasticsearch client #{client_library_version} is not compatible for your Elasticsearch server. Please check your using elasticsearch gem version and Elasticsearch server."
  end
  begin
    unless version = @_es_info.dig("version", "number")
      version = @default_elasticsearch_version
    end
  rescue NoMethodError => e
    log.warn "#{@_es_info} can not dig version information. Assuming Elasticsearch #{@default_elasticsearch_version}", error: e
    version = @default_elasticsearch_version
  end
  version.to_i
end

#dry_run? ⇒ Boolean

Returns:

  • (Boolean) 


432
433
434
435
436
437
438
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 432

def dry_run?
  if Fluent::Engine.respond_to?(:dry_run_mode)
    Fluent::Engine.dry_run_mode
  elsif Fluent::Engine.respond_to?(:supervisor_mode)
    Fluent::Engine.supervisor_mode
  end
end

#expand_placeholders(chunk) ⇒ Object 



795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 795

def expand_placeholders(chunk)
  logstash_prefix = extract_placeholders(@logstash_prefix, chunk)
  logstash_dateformat = extract_placeholders(@logstash_dateformat, chunk)
  index_name = extract_placeholders(@index_name, chunk)
  if @type_name
    type_name = extract_placeholders(@type_name, chunk)
  else
    type_name = nil
  end
  if @template_name
    template_name = extract_placeholders(@template_name, chunk)
  else
    template_name = nil
  end
  if @customize_template
    customize_template = @customize_template.each_with_object({}) { |(key, value), hash| hash[key] = extract_placeholders(value, chunk) }
  else
    customize_template = nil
  end
  if @deflector_alias
    deflector_alias = extract_placeholders(@deflector_alias, chunk)
  else
    deflector_alias = nil
  end
  if @application_name
    application_name = extract_placeholders(@application_name, chunk)
  else
    application_name = nil
  end
  if @pipeline
    pipeline = extract_placeholders(@pipeline, chunk)
  else
    pipeline = nil
  end
  if @ilm_policy_id
    ilm_policy_id = extract_placeholders(@ilm_policy_id, chunk)
  else
    ilm_policy_id = nil
  end
  return logstash_prefix, logstash_dateformat, index_name, type_name, template_name, customize_template, deflector_alias, application_name, pipeline, ilm_policy_id
end

#flatten_record(record, prefix = []) ⇒ Object 



780
781
782
783
784
785
786
787
788
789
790
791
792
793
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 780

def flatten_record(record, prefix=[])
  ret = {}
  if record.is_a? Hash
    record.each { |key, value|
      ret.merge! flatten_record(value, prefix + [key.to_s])
    }
  elsif record.is_a? Array
    # Don't mess with arrays, leave them unprocessed
    ret.merge!({prefix.join(@flatten_hashes_separator) => record})
  else
    return {prefix.join(@flatten_hashes_separator) => record}
  end
  ret
end

#get_affinity_target_indices(chunk) ⇒ Object 



920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 920

def get_affinity_target_indices(chunk)
  indices = Hash.new
  if target_index_affinity_enabled?()
    id_key_accessor = record_accessor_create(@id_key)
    ids = Set.new
    chunk.msgpack_each do |time, record|
      next unless record.is_a? Hash
      begin
        ids << id_key_accessor.call(record)
      end
    end
    log.debug("Find affinity target_indices by quering on ES (write_operation #{@write_operation}) for ids: #{ids.to_a}")
    options = {
      :index => "#{logstash_prefix}#{@logstash_prefix_separator}*",
    }
    query = {
      'query' => { 'ids' => { 'values' => ids.to_a } },
      '_source' => false,
      'sort' => [
        {"_index" => {"order" => "desc"}}
     ]
    }
    result = client.search(options.merge(:body => Yajl.dump(query)))
    # There should be just one hit per _id, but in case there still is multiple, just the oldest index is stored to map
    result['hits']['hits'].each do |hit|
      indices[hit["_id"]] = hit["_index"]
      log.debug("target_index for id: #{hit["_id"]} from es: #{hit["_index"]}")
    end
  end
  indices
end

#get_connection_options(con_host = nil) ⇒ Object 



664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 664

def get_connection_options(con_host=nil)

  hosts = if con_host || @hosts
    (con_host || @hosts).split(',').map do |host_str|
      # Support legacy hosts format host:port,host:port,host:port...
      if host_str.match(%r{^[^:]+(\:\d+)?$})
        {
          host:   host_str.split(':')[0],
          port:   (host_str.split(':')[1] || @port).to_i,
          scheme: @scheme.to_s
        }
      # Support ipv6 for host/host placeholders
      elsif is_ipv6_host(host_str)
        if Resolv::IPv6::Regex.match(host_str)
          {
            host: "[#{host_str}]",
            port: @port.to_i,
            scheme: @scheme.to_s 
          }
        else
          {
            host: host_str,
            port: @port.to_i, 
            scheme: @scheme.to_s
          }
        end
      else
        # New hosts format expects URLs such as http://logs.foo.com,https://john:[email protected]/elastic
        uri = URI(get_escaped_userinfo(host_str))
        %w(user password path).inject(host: uri.host, port: uri.port, scheme: uri.scheme) do |hash, key|
          hash[key.to_sym] = uri.public_send(key) unless uri.public_send(key).nil? || uri.public_send(key) == ''
          hash
        end
      end
    end.compact
  else
    if Resolv::IPv6::Regex.match(@host)
      [{host: "[#{@host}]", scheme: @scheme.to_s, port: @port}]
    else
      [{host: @host, port: @port, scheme: @scheme.to_s}]
    end
  end.each do |host|
    host.merge!(user: @user, password: @password) if !host[:user] && @user
    host.merge!(path: @path) if !host[:path] && @path
  end

  {
    hosts: hosts
  }
end

#get_escaped_userinfo(host_str) ⇒ Object 



644
645
646
647
648
649
650
651
652
653
654
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 644

def get_escaped_userinfo(host_str)
  if m = host_str.match(/(?<scheme>.*)%{(?<user>.*)}:%{(?<password>.*)}(?<path>@.*)/)
    m["scheme"] +
      URI.encode_www_form_component(m["user"]) +
      ':' +
      URI.encode_www_form_component(m["password"]) +
      m["path"]
  else
    host_str
  end
end

#get_parent_of(record, path) ⇒ Object

returns [parent, child_key] of child described by path array in record’s tree returns [nil, child_key] if path doesnt exist in record



1065
1066
1067
1068
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 1065

def get_parent_of(record, path)
  parent_object = path[0..-2].reduce(record) { |a, e| a.is_a?(Hash) ? a[e] : nil }
  [parent_object, path[-1]]
end

#gzip(string) ⇒ Object

gzip compress data



1071
1072
1073
1074
1075
1076
1077
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 1071

def gzip(string)
  wio = StringIO.new("w")
  w_gz = Zlib::GzipWriter.new(wio, strategy = compression_strategy)
  w_gz.write(string)
  w_gz.close
  wio.string
end

#handle_last_seen_es_major_version ⇒ Object 



484
485
486
487
488
489
490
491
492
493
494
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 484

def handle_last_seen_es_major_version
  if @verify_es_version_at_startup && !dry_run?
    retry_operate(@max_retry_get_es_version,
                  @fail_on_detecting_es_version_retry_exceed,
                  @catch_transport_exception_on_retry) do
      detect_es_major_version
    end
  else
    nil
  end
end

#inject_chunk_id_to_record_if_needed(record, chunk_id) ⇒ Object 



841
842
843
844
845
846
847
848
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 841

def inject_chunk_id_to_record_if_needed(record, chunk_id)
  if @metainfo&.include_chunk_id
    record[@metainfo.chunk_id_key] = chunk_id
    record
  else
    record
  end
end

#is_existing_connection(host) ⇒ Object 



1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 1179

def is_existing_connection(host)
  # check if the host provided match the current connection
  return false if @_es.nil?
  return false if @current_config.nil?
  return false if host.length != @current_config.length

  for i in 0...host.length
    if !host[i][:host].eql? @current_config[i][:host] || host[i][:port] != @current_config[i][:port]
      return false
    end
  end

  return true
end

#is_ipv6_host(host_str) ⇒ Object 



656
657
658
659
660
661
662
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 656

def is_ipv6_host(host_str)
  begin
    IPAddr.new(host_str).ipv6?
  rescue IPAddr::InvalidAddressError
    return false
  end
end

#multi_workers_ready? ⇒ Boolean

Returns:

  • (Boolean) 


837
838
839
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 837

def multi_workers_ready?
  true
end

#parse_time(value, event_time, tag) ⇒ Object 



575
576
577
578
579
580
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 575

def parse_time(value, event_time, tag)
  @time_parser.call(value)
rescue => e
  router.emit_error_event(@time_parse_error_tag, Fluent::Engine.now, {'tag' => tag, 'time' => event_time, 'format' => @time_key_format, 'value' => value}, e)
  return Time.at(event_time).to_datetime
end

#placeholder?(name, param) ⇒ Boolean

Returns:

  • (Boolean) 


440
441
442
443
444
445
446
447
448
449
450
451
452
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 440

def placeholder?(name, param)
  placeholder_validities = []
  placeholder_validators(name, param).each do |v|
    begin
      v.validate!
      placeholder_validities << true
    rescue Fluent::ConfigError => e
      log.debug("'#{name} #{param}' is tested built-in placeholder(s) but there is no valid placeholder(s). error: #{e}")
      placeholder_validities << false
    end
  end
  placeholder_validities.include?(true)
end

#placeholder_substitution_needed_for_template? ⇒ Boolean

Returns:

  • (Boolean) 


1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 1079

def placeholder_substitution_needed_for_template?
  need_substitution = placeholder?(:host, @host.to_s) ||
    placeholder?(:index_name, @index_name.to_s) ||
    placeholder?(:template_name, @template_name.to_s) ||
    @customize_template&.values&.any? { |value| placeholder?(:customize_template, value.to_s) } ||
    placeholder?(:logstash_prefix, @logstash_prefix.to_s) ||
    placeholder?(:logstash_dateformat, @logstash_dateformat.to_s) ||
    placeholder?(:deflector_alias, @deflector_alias.to_s) ||
    placeholder?(:application_name, @application_name.to_s) ||
    placeholder?(:ilm_policy_id, @ilm_policy_id.to_s)
  log.debug("Need substitution: #{need_substitution}")
  need_substitution
end

#process_message(tag, meta, header, time, record, affinity_target_indices, extracted_values) ⇒ Object 



964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 964

def process_message(tag, meta, header, time, record, affinity_target_indices, extracted_values)
  logstash_prefix, logstash_dateformat, index_name, type_name, _template_name, _customize_template, _deflector_alias, application_name, pipeline, _ilm_policy_id = extracted_values

  if @flatten_hashes
    record = flatten_record(record)
  end

  dt = nil
  if @logstash_format || @include_timestamp
    if record.has_key?(TIMESTAMP_FIELD)
      rts = record[TIMESTAMP_FIELD]
      dt = parse_time(rts, time, tag)
    elsif record.has_key?(@time_key)
      rts = record[@time_key]
      dt = parse_time(rts, time, tag)
      record[TIMESTAMP_FIELD] = dt.iso8601(@time_precision) unless @time_key_exclude_timestamp
    else
      dt = Time.at(time).to_datetime
      record[TIMESTAMP_FIELD] = dt.iso8601(@time_precision)
    end
  end

  target_index_parent, target_index_child_key = @target_index_key ? get_parent_of(record, @target_index_key) : nil
  if target_index_parent && target_index_parent[target_index_child_key]
    target_index_alias = target_index = target_index_parent.delete(target_index_child_key)
  elsif @logstash_format
    dt = dt.new_offset(0) if @utc_index
    target_index = "#{logstash_prefix}#{@logstash_prefix_separator}#{dt.strftime(logstash_dateformat)}"
    target_index_alias = "#{logstash_prefix}#{@logstash_prefix_separator}#{application_name}#{@logstash_prefix_separator}#{dt.strftime(logstash_dateformat)}"
  else
    target_index_alias = target_index = index_name
  end

  # Change target_index to lower-case since Elasticsearch doesn't
  # allow upper-case characters in index names.
  target_index = target_index.downcase
  target_index_alias = target_index_alias.downcase
  if @include_tag_key
    record[@tag_key] = tag
  end

  # If affinity target indices map has value for this particular id, use it as target_index
  if !affinity_target_indices.empty?
    id_accessor = record_accessor_create(@id_key)
    id_value = id_accessor.call(record)
    if affinity_target_indices.key?(id_value)
      target_index = affinity_target_indices[id_value]
    end
  end

  target_type_parent, target_type_child_key = @target_type_key ? get_parent_of(record, @target_type_key) : nil
  if target_type_parent && target_type_parent[target_type_child_key]
    target_type = target_type_parent.delete(target_type_child_key)
    if @last_seen_major_version == 6
      log.warn "Detected ES 6.x: `@type_name` will be used as the document `_type`."
      target_type = type_name
    elsif @last_seen_major_version == 7
      log.warn "Detected ES 7.x: `_doc` will be used as the document `_type`."
      target_type = '_doc'.freeze
    elsif @last_seen_major_version >= 8
      log.debug "Detected ES 8.x or above: document type will not be used."
      target_type = nil
    end
  else
    if @suppress_type_name && @last_seen_major_version == 7
      target_type = nil
    elsif @last_seen_major_version == 7 && @type_name != DEFAULT_TYPE_NAME_ES_7x
      log.warn "Detected ES 7.x: `_doc` will be used as the document `_type`."
      target_type = '_doc'.freeze
    elsif @last_seen_major_version >= 8
      log.debug "Detected ES 8.x or above: document type will not be used."
      target_type = nil
    else
      target_type = type_name
    end
  end

  meta.clear
  meta["_index".freeze] = target_index
  meta["_type".freeze] = target_type unless @last_seen_major_version >= 8
  meta["_alias".freeze] = target_index_alias

  if @pipeline
    meta["pipeline".freeze] = pipeline
  end

  @meta_config_map.each do |record_accessor, meta_key|
    if raw_value = record_accessor.call(record)
      meta[meta_key] = raw_value
    end
  end

  if @remove_keys
    @remove_keys.each { |key| record.delete(key) }
  end

  return [meta, header, record]
end

#remove_keys(record) ⇒ Object 



771
772
773
774
775
776
777
778
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 771

def remove_keys(record)
  keys = record[@remove_keys_on_update_key] || @remove_keys_on_update || []
  record.delete(@remove_keys_on_update_key)
  return record unless keys.any?
  record = record.dup
  keys.each { |key| record.delete(key) }
  record
end

#retry_stream_retryable? ⇒ Boolean

Returns:

  • (Boolean) 


1175
1176
1177
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 1175

def retry_stream_retryable?
  @buffer.storable?
end

#send_bulk(data, tag, chunk, bulk_message_count, extracted_values, info, unpacked_msg_arr) ⇒ Object

send_bulk given a specific bulk request, the original tag, chunk, and bulk_message_count



1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 1125

def send_bulk(data, tag, chunk, bulk_message_count, extracted_values, info, unpacked_msg_arr)
  _logstash_prefix, _logstash_dateformat, index_name, _type_name, template_name, customize_template, deflector_alias, application_name, _pipeline, ilm_policy_id = extracted_values
  if deflector_alias
    template_installation(deflector_alias, template_name, customize_template, application_name, index_name, ilm_policy_id, info.host)
  else
    template_installation(info.ilm_index, template_name, customize_template, application_name, @logstash_format ? info.ilm_alias : index_name, ilm_policy_id, info.host)
  end

  begin

    log.on_trace { log.trace "bulk request: #{data}" }

    prepared_data = if compression
                      gzip(data)
                    else
                      data
                    end

    response = client(info.host, compression).bulk body: prepared_data, index: info.index
    log.on_trace { log.trace "bulk response: #{response}" }

    if response['errors']
      error = Fluent::Plugin::ElasticsearchErrorHandler.new(self)
      error.handle_error(response, tag, chunk, bulk_message_count, extracted_values, unpacked_msg_arr)
    end
  rescue RetryStreamError => e
    log.trace "router.emit_stream for retry stream doing..."
    emit_tag = @retry_tag ? @retry_tag : tag
    # check capacity of buffer space
    if retry_stream_retryable?
      router.emit_stream(emit_tag, e.retry_stream)
    else
      raise RetryStreamEmitFailure, "buffer is full."
    end
    log.trace "router.emit_stream for retry stream done."
  rescue => e
    ignore = @ignore_exception_classes.any? { |clazz| e.class <= clazz }

    log.warn "Exception ignored in tag #{tag}: #{e.class.name} #{e.message}" if ignore

    @_es = nil if @reconnect_on_error
    @_es_info = nil if @reconnect_on_error

    raise UnrecoverableRequestFailure if ignore && @exception_backup

    # FIXME: identify unrecoverable errors and raise UnrecoverableRequestFailure instead
    raise RecoverableRequestFailure, "could not push logs to Elasticsearch cluster (#{connection_options_description(info.host)}): #{e.message}" unless ignore
  end
end

#setup_api_key ⇒ Object 



426
427
428
429
430
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 426

def setup_api_key
  return {} unless @api_key

  { "Authorization" => "ApiKey " + Base64.strict_encode64(@api_key) }
end

#split_request?(bulk_message, info) ⇒ Boolean

Returns:

  • (Boolean) 


952
953
954
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 952

def split_request?(bulk_message, info)
  # For safety.
end

#split_request_size_check?(bulk_message, info) ⇒ Boolean

Returns:

  • (Boolean) 


956
957
958
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 956

def split_request_size_check?(bulk_message, info)
  bulk_message[info].size > @bulk_message_request_threshold
end

#split_request_size_uncheck?(bulk_message, info) ⇒ Boolean

Returns:

  • (Boolean) 


960
961
962
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 960

def split_request_size_uncheck?(bulk_message, info)
  false
end

#target_index_affinity_enabled? ⇒ Boolean

Returns:

  • (Boolean) 


916
917
918
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 916

def target_index_affinity_enabled?()
  @target_index_affinity && @logstash_format && @id_key && (@write_operation == UPDATE_OP || @write_operation == UPSERT_OP)
end

#template_installation(deflector_alias, template_name, customize_template, application_name, ilm_policy_id, target_index, host) ⇒ Object 



1093
1094
1095
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 1093

def template_installation(deflector_alias, template_name, customize_template, application_name, ilm_policy_id, target_index, host)
  # for safety.
end

#template_installation_actual(deflector_alias, template_name, customize_template, application_name, target_index, ilm_policy_id, host = nil) ⇒ Object 



1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 1097

def template_installation_actual(deflector_alias, template_name, customize_template, application_name, target_index, ilm_policy_id, host=nil)
  if template_name && @template_file
    if !@logstash_format && (deflector_alias.nil? || (@alias_indexes.include? deflector_alias)) && (@template_names.include? template_name)
      if deflector_alias
        log.debug("Index alias #{deflector_alias} and template #{template_name} already exist (cached)")
      else
        log.debug("Template #{template_name} already exists (cached)")
      end
    else
      retry_operate(@max_retry_putting_template,
                    @fail_on_putting_template_retry_exceed,
                    @catch_transport_exception_on_retry) do
        if customize_template
          template_custom_install(template_name, @template_file, @template_overwrite, customize_template, @enable_ilm, deflector_alias, ilm_policy_id, host, target_index, @index_separator)
        else
          template_install(template_name, @template_file, @template_overwrite, @enable_ilm, deflector_alias, ilm_policy_id, host, target_index, @index_separator)
        end
        ilm_policy = @ilm_policies[ilm_policy_id] || {}
        create_rollover_alias(target_index, @rollover_index, deflector_alias, application_name, @index_date_pattern, @index_separator, @enable_ilm, ilm_policy_id, ilm_policy, @ilm_policy_overwrite, host)
      end
      @alias_indexes << deflector_alias unless deflector_alias.nil?
      @template_names << template_name
    end
  end
end

#update_body(record, op) ⇒ Object 



755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 755

def update_body(record, op)
  update = remove_keys(record)
  if @suppress_doc_wrap
    return update
  end
  body = {"doc".freeze => update}
  if op == UPSERT_OP
    if update == record
      body["doc_as_upsert".freeze] = true
    else
      body[UPSERT_OP] = record
    end
  end
  body
end

#write(chunk) ⇒ Object 



850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
 
# File 'lib/fluent/plugin/out_elasticsearch.rb', line 850

def write(chunk)
  bulk_message_count = Hash.new { |h,k| h[k] = 0 }
  bulk_message = Hash.new { |h,k| h[k] = '' }
  header = {}
  meta = {}
  unpackedMsgArr = {}

  tag = chunk.metadata.tag
  chunk_id = dump_unique_id_hex(chunk.unique_id)
  extracted_values = expand_placeholders(chunk)
  host = if @hosts
           extract_placeholders(@hosts, chunk)
         else
           extract_placeholders(@host, chunk)
         end

  affinity_target_indices = get_affinity_target_indices(chunk)
  chunk.msgpack_each do |time, record|
    next unless record.is_a? Hash

    record = inject_chunk_id_to_record_if_needed(record, chunk_id)

    begin
      meta, header, record = process_message(tag, meta, header, time, record, affinity_target_indices, extracted_values)
      info = if @include_index_in_url
               RequestInfo.new(host, meta.delete("_index".freeze), meta["_index".freeze], meta.delete("_alias".freeze))
             else
               RequestInfo.new(host, nil, meta["_index".freeze], meta.delete("_alias".freeze))
             end

      unpackedMsgArr[info] = [] if unpackedMsgArr[info].nil?
      unpackedMsgArr[info] << {:time => time, :record => record}

      if split_request?(bulk_message, info)
        bulk_message.each do |info, msgs|
          send_bulk(msgs, tag, chunk, bulk_message_count[info], extracted_values, info, unpackedMsgArr[info]) unless msgs.empty?
          unpackedMsgArr[info].clear
          msgs.clear
          # Clear bulk_message_count for this info.
          bulk_message_count[info] = 0;
          next
        end
      end

      if append_record_to_messages(@write_operation, meta, header, record, bulk_message[info])
        bulk_message_count[info] += 1;
      else
        if @emit_error_for_missing_id
          raise MissingIdFieldError, "Missing '_id' field. Write operation is #{@write_operation}"
        else
          log.on_debug { log.debug("Dropping record because its missing an '_id' field and write_operation is #{@write_operation}: #{record}") }
        end
      end
    rescue => e
      router.emit_error_event(tag, time, record, e)
    end
  end

  bulk_message.each do |info, msgs|
    send_bulk(msgs, tag, chunk, bulk_message_count[info], extracted_values, info, unpackedMsgArr[info]) unless msgs.empty?

    unpackedMsgArr[info].clear
    msgs.clear
  end
end