Class: Fluent::KafkaOutput

Inherits:
Output
  • Object
show all
Includes:
Fluent::KafkaPluginUtil::SSLSettings, Fluent::KafkaPluginUtil::SaslSettings
Defined in:
lib/fluent/plugin/out_kafka.rb

Constant Summary

Constants included from Fluent::KafkaPluginUtil::SSLSettings

Fluent::KafkaPluginUtil::SSLSettings::DummyFormatter

Instance Attribute Summary collapse

Instance Method Summary collapse

Methods included from Fluent::KafkaPluginUtil::SaslSettings

included

Methods included from Fluent::KafkaPluginUtil::SSLSettings

included, #pickup_ssl_endpoint, #read_ssl_file

Constructor Details

#initialize ⇒ KafkaOutput

Returns a new instance of KafkaOutput.



85
86
87
88
89
90
91
 
# File 'lib/fluent/plugin/out_kafka.rb', line 85

def initialize
  super

  require 'kafka'

  @kafka = nil
end

Instance Attribute Details

#field_separator ⇒ Object

Returns the value of attribute field_separator.



79
80
81
 
# File 'lib/fluent/plugin/out_kafka.rb', line 79

def field_separator
  @field_separator
end

#output_data_type ⇒ Object

Returns the value of attribute output_data_type.



78
79
80
 
# File 'lib/fluent/plugin/out_kafka.rb', line 78

def output_data_type
  @output_data_type
end

Instance Method Details

#configure(conf) ⇒ Object 



136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
 
# File 'lib/fluent/plugin/out_kafka.rb', line 136

def configure(conf)
  super

  log.warn "Support of fluentd v0.12 has ended. Use kafka2 instead. kafka will be an alias of kafka2"

  if @zookeeper
    require 'zookeeper'
  else
    @seed_brokers = @brokers.split(",")
    log.info "brokers has been set directly: #{@seed_brokers}"
  end

  if conf['ack_timeout_ms']
    log.warn "'ack_timeout_ms' parameter is deprecated. Use second unit 'ack_timeout' instead"
    @ack_timeout = conf['ack_timeout_ms'].to_i / 1000
  end

  @f_separator = case @field_separator
                 when /SPACE/i then ' '
                 when /COMMA/i then ','
                 when /SOH/i then "\x01"
                 else "\t"
                 end

  @formatter_proc = setup_formatter(conf)

  @producer_opts = {max_retries: @max_send_retries, required_acks: @required_acks}
  @producer_opts[:ack_timeout] = @ack_timeout if @ack_timeout
  @producer_opts[:compression_codec] = @compression_codec.to_sym if @compression_codec
  @producer_opts[:max_buffer_size] = @max_buffer_size if @max_buffer_size
  @producer_opts[:max_buffer_bytesize] = @max_buffer_bytesize if @max_buffer_bytesize
  if @active_support_notification_regex
    require 'active_support/notifications'
    require 'active_support/core_ext/hash/keys'
    ActiveSupport::Notifications.subscribe(Regexp.new(@active_support_notification_regex)) do |*args|
      event = ActiveSupport::Notifications::Event.new(*args)
      message = event.payload.respond_to?(:stringify_keys) ? event.payload.stringify_keys : event.payload
      @router.emit("fluent_kafka_stats.#{event.name}", Time.now.to_i, message)
    end
  end
end

#emit(tag, es, chain) ⇒ Object 



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
 
# File 'lib/fluent/plugin/out_kafka.rb', line 218

def emit(tag, es, chain)
  begin
    chain.next

    # out_kafka is mainly for testing so don't need the performance unlike out_kafka_buffered.
    producer = @kafka.producer(**@producer_opts)

    es.each do |time, record|
      if @output_include_time
        if @time_format
          record['time'] = Time.at(time).strftime(@time_format)
        else
          record['time'] = time
        end
      end
      record['tag'] = tag if @output_include_tag
      topic = (@exclude_topic_key ? record.delete('topic') : record['topic']) || @default_topic || tag
      partition_key = (@exclude_partition_key ? record.delete('partition_key') : record['partition_key']) || @default_partition_key
      partition = (@exclude_partition ? record.delete('partition'.freeze) : record['partition'.freeze]) || @default_partition
      message_key = (@exclude_message_key ? record.delete('message_key') : record['message_key']) || @default_message_key

      record_buf = @formatter_proc.call(tag, time, record)
      record_buf_bytes = record_buf.bytesize
      if @max_send_limit_bytes && record_buf_bytes > @max_send_limit_bytes
        log.warn "record size exceeds max_send_limit_bytes. Skip event:", :time => time, :record => record
        next
      end
      log.trace { "message will send to #{topic} with partition_key: #{partition_key}, partition: #{partition}, message_key: #{message_key} and value: #{record_buf}." }
      begin
        producer.produce(record_buf, topic: topic, key: message_key, partition: partition, partition_key: partition_key)
      rescue Kafka::BufferOverflow => e
        log.warn "BufferOverflow occurred: #{e}"
        log.info "Trying to deliver the messages to prevent the buffer from overflowing again."
        producer.deliver_messages
        log.info "Recovered from BufferOverflow successfully`"
      end
    end

    producer.deliver_messages
    producer.shutdown
  rescue Exception => e
    log.warn "Send exception occurred: #{e}"
    producer.shutdown if producer
    refresh_client
    raise e
  end
end

#multi_workers_ready? ⇒ Boolean

Returns:

  • (Boolean) 


178
179
180
 
# File 'lib/fluent/plugin/out_kafka.rb', line 178

def multi_workers_ready?
  true
end

#refresh_client ⇒ Object 



93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
 
# File 'lib/fluent/plugin/out_kafka.rb', line 93

def refresh_client
  if @zookeeper
    @seed_brokers = []
    z = Zookeeper.new(@zookeeper)
    z.get_children(:path => @zookeeper_path)[:children].each do |id|
      broker = Yajl.load(z.get(:path => @zookeeper_path + "/#{id}")[:data])
      if @ssl_client_cert
        @seed_brokers.push(pickup_ssl_endpoint(broker))
      else
        @seed_brokers.push("#{broker['host']}:#{broker['port']}")
      end
    end
    z.close
    log.info "brokers has been refreshed via Zookeeper: #{@seed_brokers}"
  end
  begin
    if @seed_brokers.length > 0
      if @scram_mechanism != nil && @username != nil && @password != nil
        @kafka = Kafka.new(seed_brokers: @seed_brokers, client_id: @client_id, ssl_ca_cert_file_path: @ssl_ca_cert,
                           ssl_client_cert: read_ssl_file(@ssl_client_cert), ssl_client_cert_key: read_ssl_file(@ssl_client_cert_key), ssl_ca_certs_from_system: @ssl_ca_certs_from_system,
                           sasl_scram_username: @username, sasl_scram_password: @password, sasl_scram_mechanism: @scram_mechanism, sasl_over_ssl: @sasl_over_ssl,
                           ssl_verify_hostname: @ssl_verify_hostname,
                           partitioner: Kafka::Partitioner.new(hash_function: @partitioner_hash_function))
      elsif @username != nil && @password != nil
        @kafka = Kafka.new(seed_brokers: @seed_brokers, client_id: @client_id, ssl_ca_cert_file_path: @ssl_ca_cert,
                           ssl_client_cert: read_ssl_file(@ssl_client_cert), ssl_client_cert_key: read_ssl_file(@ssl_client_cert_key), ssl_ca_certs_from_system: @ssl_ca_certs_from_system,
                           sasl_plain_username: @username, sasl_plain_password: @password, sasl_over_ssl: @sasl_over_ssl, ssl_verify_hostname: @ssl_verify_hostname,
                           partitioner: Kafka::Partitioner.new(hash_function: @partitioner_hash_function))
      else
        @kafka = Kafka.new(seed_brokers: @seed_brokers, client_id: @client_id, ssl_ca_cert_file_path: @ssl_ca_cert,
                           ssl_client_cert: read_ssl_file(@ssl_client_cert), ssl_client_cert_key: read_ssl_file(@ssl_client_cert_key), ssl_ca_certs_from_system: @ssl_ca_certs_from_system,
                           sasl_gssapi_principal: @principal, sasl_gssapi_keytab: @keytab, sasl_over_ssl: @sasl_over_ssl, ssl_verify_hostname: @ssl_verify_hostname,
                           partitioner: Kafka::Partitioner.new(hash_function: @partitioner_hash_function))
      end
      log.info "initialized kafka producer: #{@client_id}"
    else
      log.warn "No brokers found on Zookeeper"
    end
  rescue Exception => e
    log.error e
  end
end

#setup_formatter(conf) ⇒ Object 



192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
 
# File 'lib/fluent/plugin/out_kafka.rb', line 192

def setup_formatter(conf)
  if @output_data_type == 'json'
    require 'yajl'
    Proc.new { |tag, time, record| Yajl::Encoder.encode(record) }
  elsif @output_data_type == 'ltsv'
    require 'ltsv'
    Proc.new { |tag, time, record| LTSV.dump(record) }
  elsif @output_data_type == 'msgpack'
    require 'msgpack'
    Proc.new { |tag, time, record| record.to_msgpack }
  elsif @output_data_type =~ /^attr:(.*)$/
    @custom_attributes = $1.split(',').map(&:strip).reject(&:empty?)
    @custom_attributes.unshift('time') if @output_include_time
    @custom_attributes.unshift('tag') if @output_include_tag
    Proc.new { |tag, time, record|
      @custom_attributes.map { |attr|
        record[attr].nil? ? '' : record[attr].to_s
      }.join(@f_separator)
    }
  else
    @formatter = Fluent::Plugin.new_formatter(@output_data_type)
    @formatter.configure(conf)
    @formatter.method(:format)
  end
end

#shutdown ⇒ Object 



187
188
189
190
 
# File 'lib/fluent/plugin/out_kafka.rb', line 187

def shutdown
  super
  @kafka = nil
end

#start ⇒ Object 



182
183
184
185
 
# File 'lib/fluent/plugin/out_kafka.rb', line 182

def start
  super
  refresh_client
end