Class: LogStash::JavaPipeline

Inherits:
JavaBasePipeline
  • Object
show all
Includes:
Util::Loggable
Defined in:
lib/logstash/java_pipeline.rb

Constant Summary collapse

MAX_INFLIGHT_WARN_THRESHOLD = 
10_000

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(pipeline_config, namespaced_metric = nil, agent = nil) ⇒ JavaPipeline

Returns a new instance of JavaPipeline.



22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
 
# File 'lib/logstash/java_pipeline.rb', line 22

def initialize(pipeline_config, namespaced_metric = nil, agent = nil)
  @logger = self.logger
  super pipeline_config, namespaced_metric, @logger, agent
  open_queue

  @worker_threads = []

  @drain_queue =  settings.get_value("queue.drain") || settings.get("queue.type") == "memory"

  @events_filtered = java.util.concurrent.atomic.LongAdder.new
  @events_consumed = java.util.concurrent.atomic.LongAdder.new

  @input_threads = []
  # @ready requires thread safety since it is typically polled from outside the pipeline thread
  @ready = Concurrent::AtomicBoolean.new(false)
  @running = Concurrent::AtomicBoolean.new(false)
  @flushing = java.util.concurrent.atomic.AtomicBoolean.new(false)
  @flushRequested = java.util.concurrent.atomic.AtomicBoolean.new(false)
  @shutdownRequested = java.util.concurrent.atomic.AtomicBoolean.new(false)
  @outputs_registered = Concurrent::AtomicBoolean.new(false)
  @finished_execution = Concurrent::AtomicBoolean.new(false)
end

Instance Attribute Details

#events_consumed ⇒ Object (readonly)

Returns the value of attribute events_consumed.



13
14
15
 
# File 'lib/logstash/java_pipeline.rb', line 13

def events_consumed
  @events_consumed
end

#events_filtered ⇒ Object (readonly)

Returns the value of attribute events_filtered.



13
14
15
 
# File 'lib/logstash/java_pipeline.rb', line 13

def events_filtered
  @events_filtered
end

#started_at ⇒ Object (readonly)

Returns the value of attribute started_at.



13
14
15
 
# File 'lib/logstash/java_pipeline.rb', line 13

def started_at
  @started_at
end

#thread ⇒ Object (readonly)

Returns the value of attribute thread.



13
14
15
 
# File 'lib/logstash/java_pipeline.rb', line 13

def thread
  @thread
end

#worker_threads ⇒ Object (readonly)

Returns the value of attribute worker_threads.



13
14
15
 
# File 'lib/logstash/java_pipeline.rb', line 13

def worker_threads
  @worker_threads
end

Instance Method Details

#clear_pipeline_metrics ⇒ Object 



389
390
391
392
393
394
395
396
397
398
399
400
401
 
# File 'lib/logstash/java_pipeline.rb', line 389

def clear_pipeline_metrics
  # TODO(ph): I think the metric should also proxy that call correctly to the collector
  # this will simplify everything since the null metric would simply just do a noop
  collector = metric.collector

  unless collector.nil?
    # selectively reset metrics we don't wish to keep after reloading
    # these include metrics about the plugins and number of processed events
    # we want to keep other metrics like reload counts and error messages
    collector.clear("stats/pipelines/#{pipeline_id}/plugins")
    collector.clear("stats/pipelines/#{pipeline_id}/events")
  end
end

#filter(event, &block) ⇒ Object

for backward compatibility in devutils for the rspec helpers, this method is not used anymore and just here to not break TestPipeline that inherits this class.



348
349
 
# File 'lib/logstash/java_pipeline.rb', line 348

def filter(event, &block)
end

#filters? ⇒ Boolean

Returns:

  • (Boolean) 


73
74
75
 
# File 'lib/logstash/java_pipeline.rb', line 73

def filters?
  filters.any?
end

#flush_filters(options = {}, &block) ⇒ Object

for backward compatibility in devutils for the rspec helpers, this method is not used anymore and just here to not break TestPipeline that inherits this class.



353
354
 
# File 'lib/logstash/java_pipeline.rb', line 353

def flush_filters(options = {}, &block)
end

#inputworker(plugin) ⇒ Object 



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
 
# File 'lib/logstash/java_pipeline.rb', line 269

def inputworker(plugin)
  Util::set_thread_name("[#{pipeline_id}]<#{plugin.class.config_name}")
  begin
    plugin.run(wrapped_write_client(plugin.id.to_sym))
  rescue => e
    if plugin.stop?
      @logger.debug("Input plugin raised exception during shutdown, ignoring it.",
                    default_logging_keys(:plugin => plugin.class.config_name, :exception => e.message, :backtrace => e.backtrace))
      return
    end

    # otherwise, report error and restart
    @logger.error(I18n.t("logstash.pipeline.worker-error-debug",
                          default_logging_keys(
                            :plugin => plugin.inspect,
                            :error => e.message,
                            :exception => e.class,
                            :stacktrace => e.backtrace.join("\n"))))

    # Assuming the failure that caused this exception is transient,
    # let's sleep for a bit and execute #run again
    sleep(1)
    retry
  ensure
    plugin.do_close
  end
end

#inspect ⇒ Object

Sometimes we log stuff that will dump the pipeline which may contain sensitive information (like the raw syntax tree which can contain passwords) We want to hide most of what’s in here



406
407
408
409
410
411
412
413
414
 
# File 'lib/logstash/java_pipeline.rb', line 406

def inspect
  {
    :pipeline_id => pipeline_id,
    :settings => settings.inspect,
    :ready => @ready,
    :running => @running,
    :flushing => @flushing
  }
end

#plugin_threads_info ⇒ Object 



375
376
377
378
379
 
# File 'lib/logstash/java_pipeline.rb', line 375

def plugin_threads_info
  input_threads = @input_threads.select {|t| t.alive? }
  worker_threads = @worker_threads.select {|t| t.alive? }
  (input_threads + worker_threads).map {|t| Util.thread_info(t) }
end

#ready? ⇒ Boolean

def initialize

Returns:

  • (Boolean) 


45
46
47
 
# File 'lib/logstash/java_pipeline.rb', line 45

def ready?
  @ready.value
end

#register_plugins(plugins) ⇒ Object

register_plugins calls #register_plugin on the plugins list and upon exception will call Plugin#do_close on all registered plugins

Parameters:

  • plugins (Array[Plugin]) —

    the list of plugins to register



173
174
175
176
177
178
179
180
181
182
 
# File 'lib/logstash/java_pipeline.rb', line 173

def register_plugins(plugins)
  registered = []
  plugins.each do |plugin|
    plugin.register
    registered << plugin
  end
rescue => e
  registered.each(&:do_close)
  raise e
end

#run ⇒ Object 



125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
 
# File 'lib/logstash/java_pipeline.rb', line 125

def run
  @started_at = Time.now
  @thread = Thread.current
  Util.set_thread_name("[#{pipeline_id}]-pipeline-manager")

  start_workers

  @logger.info("Pipeline started", "pipeline.id" => pipeline_id)

  # Block until all inputs have stopped
  # Generally this happens if SIGINT is sent and `shutdown` is called from an external thread

  transition_to_running
  start_flusher # Launches a non-blocking thread for flush events
  wait_inputs
  transition_to_stopped

  @logger.debug("Input plugins stopped! Will shutdown filter/output workers.", default_logging_keys)

  shutdown_flusher
  shutdown_workers

  close

  @logger.debug("Pipeline has been shutdown", default_logging_keys)

  # exit code
  return 0
end

#running? ⇒ Boolean

Returns:

  • (Boolean) 


163
164
165
 
# File 'lib/logstash/java_pipeline.rb', line 163

def running?
  @running.true?
end

#safe_pipeline_worker_count ⇒ Object 



49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
 
# File 'lib/logstash/java_pipeline.rb', line 49

def safe_pipeline_worker_count
  default = settings.get_default("pipeline.workers")
  pipeline_workers = settings.get("pipeline.workers") #override from args "-w 8" or config
  safe_filters, unsafe_filters = filters.partition(&:threadsafe?)
  plugins = unsafe_filters.collect { |f| f.config_name }

  return pipeline_workers if unsafe_filters.empty?

  if settings.set?("pipeline.workers")
    if pipeline_workers > 1
      @logger.warn("Warning: Manual override - there are filters that might not work with multiple worker threads", default_logging_keys(:worker_threads => pipeline_workers, :filters => plugins))
    end
  else
    # user did not specify a worker thread count
    # warn if the default is multiple
    if default > 1
      @logger.warn("Defaulting pipeline worker threads to 1 because there are some filters that might not work with multiple worker threads",
                   default_logging_keys(:count_was => default, :filters => plugins))
      return 1 # can't allow the default value to propagate if there are unsafe filters
    end
  end
  pipeline_workers
end

#shutdown(&before_stop) ⇒ Object

initiate the pipeline shutdown sequence this method is intended to be called from outside the pipeline thread

Parameters:

  • before_stop (Proc) —

    code block called before performing stop operation on input plugins



300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
 
# File 'lib/logstash/java_pipeline.rb', line 300

def shutdown(&before_stop)
  # shutdown can only start once the pipeline has completed its startup.
  # avoid potential race condition between the startup sequence and this
  # shutdown method which can be called from another thread at any time
  sleep(0.1) while !ready?

  # TODO: should we also check against calling shutdown multiple times concurrently?

  before_stop.call if block_given?

  stop_inputs

  # We make this call blocking, so we know for sure when the method return the shtudown is
  # stopped
  wait_for_workers
  clear_pipeline_metrics
  @logger.info("Pipeline terminated", "pipeline.id" => pipeline_id)
end

#shutdown_flusher ⇒ Object 



363
364
365
 
# File 'lib/logstash/java_pipeline.rb', line 363

def shutdown_flusher
  @flusher_thread.close
end

#shutdown_workers ⇒ Object

After ‘shutdown` is called from an external thread this is called from the main thread to tell the worker threads to stop and then block until they’ve fully stopped This also stops all filter and output plugins



334
335
336
337
338
339
340
341
342
343
344
 
# File 'lib/logstash/java_pipeline.rb', line 334

def shutdown_workers
  @shutdownRequested.set(true)

  @worker_threads.each do |t|
    @logger.debug("Shutdown waiting for worker thread" , default_logging_keys(:thread => t.inspect))
    t.join
  end

  filters.each(&:do_close)
  outputs.each(&:do_close)
end

#stalling_threads_info ⇒ Object 



381
382
383
384
385
386
387
 
# File 'lib/logstash/java_pipeline.rb', line 381

def stalling_threads_info
  plugin_threads_info
    .reject {|t| t["blocked_on"] } # known benign blocking statuses
    .each {|t| t.delete("backtrace") }
    .each {|t| t.delete("blocked_on") }
    .each {|t| t.delete("status") }
end

#start ⇒ Object 



77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
 
# File 'lib/logstash/java_pipeline.rb', line 77

def start
  # Since we start lets assume that the metric namespace is cleared
  # this is useful in the context of pipeline reloading
  collect_stats
  collect_dlq_stats

  @logger.debug("Starting pipeline", default_logging_keys)

  @finished_execution.make_false

  @thread = Thread.new do
    begin
      LogStash::Util.set_thread_name("pipeline.#{pipeline_id}")
      run
      @finished_execution.make_true
    rescue => e
      close
      logger.error("Pipeline aborted due to error", default_logging_keys(:exception => e, :backtrace => e.backtrace))
    end
  end

  status = wait_until_started

  if status
    logger.debug("Pipeline started successfully", default_logging_keys(:pipeline_id => pipeline_id))
  end

  status
end

#start_flusher ⇒ Object 



356
357
358
359
360
361
 
# File 'lib/logstash/java_pipeline.rb', line 356

def start_flusher
  # Invariant to help detect improper initialization
  raise "Attempted to start flusher on a stopped pipeline!" if stopped?
  @flusher_thread = org.logstash.execution.PeriodicFlush.new(@flushRequested, @flushing)
  @flusher_thread.start
end

#start_input(plugin) ⇒ Object 



265
266
267
 
# File 'lib/logstash/java_pipeline.rb', line 265

def start_input(plugin)
  @input_threads << Thread.new { inputworker(plugin) }
end

#start_inputs ⇒ Object 



247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
 
# File 'lib/logstash/java_pipeline.rb', line 247

def start_inputs
  moreinputs = []
  inputs.each do |input|
    if input.threadable && input.threads > 1
      (input.threads - 1).times do |i|
        moreinputs << input.clone
      end
    end
  end
  moreinputs.each {|i| inputs << i}

  # first make sure we can register all input plugins
  register_plugins(inputs)

  # then after all input plugins are successfully registered, start them
  inputs.each { |input| start_input(input) }
end

#start_workers ⇒ Object 



184
185
186
187
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
236
237
238
239
240
241
 
# File 'lib/logstash/java_pipeline.rb', line 184

def start_workers
  @worker_threads.clear # In case we're restarting the pipeline
  @outputs_registered.make_false
  begin
    maybe_setup_out_plugins

    pipeline_workers = safe_pipeline_worker_count
    batch_size = settings.get("pipeline.batch.size")
    batch_delay = settings.get("pipeline.batch.delay")

    max_inflight = batch_size * pipeline_workers

    config_metric = metric.namespace([:stats, :pipelines, pipeline_id.to_s.to_sym, :config])
    config_metric.gauge(:workers, pipeline_workers)
    config_metric.gauge(:batch_size, batch_size)
    config_metric.gauge(:batch_delay, batch_delay)
    config_metric.gauge(:config_reload_automatic, settings.get("config.reload.automatic"))
    config_metric.gauge(:config_reload_interval, settings.get("config.reload.interval"))
    config_metric.gauge(:dead_letter_queue_enabled, dlq_enabled?)
    config_metric.gauge(:dead_letter_queue_path, dlq_writer.get_path.to_absolute_path.to_s) if dlq_enabled?


    @logger.info("Starting pipeline", default_logging_keys(
      "pipeline.workers" => pipeline_workers,
      "pipeline.batch.size" => batch_size,
      "pipeline.batch.delay" => batch_delay,
      "pipeline.max_inflight" => max_inflight))
    if max_inflight > MAX_INFLIGHT_WARN_THRESHOLD
      @logger.warn("CAUTION: Recommended inflight events max exceeded! Logstash will run with up to #{max_inflight} events in memory in your current configuration. If your message sizes are large this may cause instability with the default heap size. Please consider setting a non-standard heap size, changing the batch size (currently #{batch_size}), or changing the number of pipeline workers (currently #{pipeline_workers})", default_logging_keys)
    end

    filter_queue_client.set_batch_dimensions(batch_size, batch_delay)

    pipeline_workers.times do |t|
      thread = Thread.new do
        org.logstash.execution.WorkerLoop.new(
            lir_execution, filter_queue_client, @events_filtered, @events_consumed,
            @flushRequested, @flushing, @shutdownRequested, @drain_queue).run
      end
      Util.set_thread_name("[#{pipeline_id}]>worker#{t}")
      @worker_threads << thread
    end

    # inputs should be started last, after all workers
    begin
      start_inputs
    rescue => e
      # if there is any exception in starting inputs, make sure we shutdown workers.
      # exception will already by logged in start_inputs
      shutdown_workers
      raise e
    end
  ensure
    # it is important to guarantee @ready to be true after the startup sequence has been completed
    # to potentially unblock the shutdown method which may be waiting on @ready to proceed
    @ready.make_true
  end
end

#stop_inputs ⇒ Object 



325
326
327
328
329
 
# File 'lib/logstash/java_pipeline.rb', line 325

def stop_inputs
  @logger.debug("Closing inputs", default_logging_keys)
  inputs.each(&:do_stop)
  @logger.debug("Closed inputs", default_logging_keys)
end

#stopped? ⇒ Boolean

Returns:

  • (Boolean) 


167
168
169
 
# File 'lib/logstash/java_pipeline.rb', line 167

def stopped?
  @running.false?
end

#transition_to_running ⇒ Object

def run



155
156
157
 
# File 'lib/logstash/java_pipeline.rb', line 155

def transition_to_running
  @running.make_true
end

#transition_to_stopped ⇒ Object 



159
160
161
 
# File 'lib/logstash/java_pipeline.rb', line 159

def transition_to_stopped
  @running.make_false
end

#uptime ⇒ Fixnum

Calculate the uptime in milliseconds

Returns:

  • (Fixnum) —

    Uptime in milliseconds, 0 if the pipeline is not started



370
371
372
373
 
# File 'lib/logstash/java_pipeline.rb', line 370

def uptime
  return 0 if started_at.nil?
  ((Time.now.to_f - started_at.to_f) * 1000.0).to_i
end

#wait_for_workers ⇒ Object

def shutdown



319
320
321
322
323
 
# File 'lib/logstash/java_pipeline.rb', line 319

def wait_for_workers
  @logger.debug("Closing inputs", default_logging_keys)
  @worker_threads.map(&:join)
  @logger.debug("Worker closed", default_logging_keys)
end

#wait_inputs ⇒ Object 



243
244
245
 
# File 'lib/logstash/java_pipeline.rb', line 243

def wait_inputs
  @input_threads.each(&:join)
end

#wait_until_started ⇒ Object 



107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
 
# File 'lib/logstash/java_pipeline.rb', line 107

def wait_until_started
  while true do
    # This should be changed with an appropriate FSM
    # It's an edge case, if we have a pipeline with
    # a generator { count => 1 } its possible that `Thread#alive?` doesn't return true
    # because the execution of the thread was successful and complete
    if @finished_execution.true?
      return true
    elsif thread.nil? || !thread.alive?
      return false
    elsif running?
      return true
    else
      sleep 0.01
    end
  end
end