Class: Fluent::KafkaOutputBuffered

Inherits:
BufferedOutput
  • Object
show all
Includes:
Fluent::KafkaPluginUtil::SSLSettings
Defined in:
lib/fluent/plugin/out_kafka_buffered.rb

Instance Attribute Summary collapse

Instance Method Summary collapse

Methods included from Fluent::KafkaPluginUtil::SSLSettings

included, #read_ssl_file

Constructor Details

#initialize ⇒ KafkaOutputBuffered

Returns a new instance of KafkaOutputBuffered.



80
81
82
83
84
85
86
87
88
89
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 80

def initialize
  super

  require 'kafka'
  require 'fluent/plugin/kafka_producer_ext'

  @kafka = nil
  @producers = {}
  @producers_mutex = Mutex.new
end

Instance Attribute Details

#field_separator ⇒ Object

Returns the value of attribute field_separator.



74
75
76
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 74

def field_separator
  @field_separator
end

#output_data_type ⇒ Object

Returns the value of attribute output_data_type.



73
74
75
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 73

def output_data_type
  @output_data_type
end

Instance Method Details

#configure(conf) ⇒ Object 



120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 120

def configure(conf)
  super

  if @zookeeper
    require 'zookeeper'
  else
    @seed_brokers = @brokers.match(",").nil? ? [@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
end

#emit(tag, es, chain) ⇒ Object 



160
161
162
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 160

def emit(tag, es, chain)
  super(tag, es, chain, tag)
end

#format_stream(tag, es) ⇒ Object 



164
165
166
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 164

def format_stream(tag, es)
  es.to_msgpack_stream
end

#get_producer ⇒ Object 



177
178
179
180
181
182
183
184
185
186
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 177

def get_producer
  @producers_mutex.synchronize {
    producer = @producers[Thread.current.object_id]
    unless producer
      producer = @kafka.producer(@producer_opts)
      @producers[Thread.current.object_id] = producer
    end
    producer
  }
end

#refresh_client(raise_error = true) ⇒ Object 



91
92
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
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 91

def refresh_client(raise_error = true)
  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])
      @seed_brokers.push("#{broker['host']}:#{broker['port']}")
    end
    z.close
    log.info "brokers has been refreshed via Zookeeper: #{@seed_brokers}"
  end
  begin
    if @seed_brokers.length > 0
      logger = @get_kafka_client_log ? log : nil
      @kafka = Kafka.new(seed_brokers: @seed_brokers, client_id: @client_id, logger: logger, ssl_ca_cert: read_ssl_file(@ssl_ca_cert),
                         ssl_client_cert: read_ssl_file(@ssl_client_cert), ssl_client_cert_key: read_ssl_file(@ssl_client_cert_key))
      log.info "initialized kafka producer: #{@client_id}"
    else
      log.warn "No brokers found on Zookeeper"
    end
  rescue Exception => e
    if raise_error # During startup, error should be reported to engine and stop its phase for safety.
      raise e
    else
      log.error e
    end
  end
end

#setup_formatter(conf) ⇒ Object 



188
189
190
191
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
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 188

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 == 'avro'
    require "avro_turf"
    require 'avro_turf/messaging'
    require "avro/builder"
    Proc.new do |tag, time, record|
      record = record.select{|key, value| !key.nil? && !key.empty?}.map{|k, v| [k.tr('[]-', '_').delete('$'), ((!v.is_a?(Fixnum) && !v.is_a?(Float)) ? v.to_s.force_encoding("UTF-8") : v)]}.to_h
      timestamp = Time.new
      record['enchilada_time_with_format'] = timestamp.strftime("%Y-%m-%dT%H:%M:%S.%LZ")

      fields = record.map{|key, value| {'name' => key, 'type' => (value.is_a?(Fixnum) ? 'int' : (value.is_a?(Float) ? 'float' : 'string'))}}
      record['enchilada_timestamp'] = timestamp.strftime('%s%3N').to_i
      fields << {"name" => "enchilada_timestamp", "type" => "long"}
      @topic_name = schema_name = "#{tag.to_s.tr('.$:', '_')}_#{Digest::MD5.new.hexdigest(fields.to_s)[0..5]}"
      schema_json = {
          "type": "record",
          "name": schema_name,
          "fields": fields
      }.to_json
      schema = Avro::Schema.parse(schema_json)

      avro = AvroTurf::Messaging.new(registry_url: @schema_registry)
      avro.encode(record, schema: schema, subject: "#{schema_name}-value")
    end
  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 



154
155
156
157
158
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 154

def shutdown
  super
  shutdown_producers
  @kafka = nil
end

#shutdown_producers ⇒ Object 



168
169
170
171
172
173
174
175
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 168

def shutdown_producers
  @producers_mutex.synchronize {
    @producers.each { |key, producer|
      producer.shutdown
    }
    @producers = {}
  }
end

#start ⇒ Object 



149
150
151
152
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 149

def start
  super
  refresh_client
end

#write(chunk) ⇒ Object 



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
 
# File 'lib/fluent/plugin/out_kafka_buffered.rb', line 237

def write(chunk)
  tag = chunk.key
  producer = get_producer

  records_by_topic = {}
  bytes_by_topic = {}
  messages = 0
  messages_bytes = 0
  record_buf = nil
  record_buf_bytes = nil

  begin
    chunk.msgpack_each { |time, record|
      begin
        if @output_include_time
          if @time_format
            record['time'.freeze] = Time.at(time).strftime(@time_format)
          else
            record['time'.freeze] = time
          end
        end

        record['tag'] = tag if @output_include_tag
        record_buf = @formatter_proc.call(tag, time, record)
        topic = (@exclude_topic_key ? record.delete('topic'.freeze) : record['topic'.freeze]) || @topic_name
        partition_key = (@exclude_partition_key ? record.delete('partition_key'.freeze) : record['partition_key'.freeze]) || @default_partition_key
        partition = (@exclude_partition ? record.delete('partition'.freeze) : record['partition'.freeze]) || @default_partition
        message_key = (@exclude_message_key ? record.delete('message_key'.freeze) : record['message_key'.freeze]) || @default_message_key

        records_by_topic[topic] ||= 0
        bytes_by_topic[topic] ||= 0

        record_buf = @formatter_proc.call(tag, time, record)
        record_buf_bytes = record_buf.bytesize
      rescue StandardError => e
        log.warn "unexpected error during format record. Skip broken event:", :error => e.to_s, :error_class => e.class.to_s, :time => time, :record => record
        next
      end

      if (messages > 0) and (messages_bytes + record_buf_bytes > @kafka_agg_max_bytes)
        log.on_trace { log.trace("#{messages} messages send.") }
        producer.deliver_messages
        messages = 0
        messages_bytes = 0
      end
      log.on_trace { log.trace("message will send to #{topic} with partition_key: #{partition_key}, partition: #{partition}, message_key: #{message_key} and value: #{record_buf}.") }
      messages += 1
      producer.produce2(record_buf, topic: topic, key: message_key, partition_key: partition_key, partition: partition)
      messages_bytes += record_buf_bytes

      records_by_topic[topic] += 1
      bytes_by_topic[topic] += record_buf_bytes
    }
    if messages > 0
      log.trace { "#{messages} messages send." }
      producer.deliver_messages
    end
    log.debug { "(records|bytes) (#{records_by_topic}|#{bytes_by_topic})" }
  end
rescue Exception => e
  log.warn "Send exception occurred: #{e}"
  log.warn "Exception Backtrace : #{e.backtrace.join("\n")}"
  # For safety, refresh client and its producers
  shutdown_producers
  refresh_client(false)
  # Raise exception to retry sendind messages
  raise e
end