Class: Bricolage::StreamingLoad::Dispatcher

Inherits:
Bricolage::SQSDataSource::MessageHandler show all
Defined in:
lib/bricolage/streamingload/dispatcher.rb

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from Bricolage::SQSDataSource::MessageHandler

#call_handler_method, #handle, #handleable?, #handler_method

Constructor Details

#initialize(event_queue:, task_queue:, chunk_router:, chunk_buffer:, task_logger:, dispatch_interval:, logger:) ⇒ Dispatcher

Returns a new instance of Dispatcher.



105
106
107
108
109
110
111
112
113
114
115
116
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 105

def initialize(event_queue:, task_queue:, chunk_router:, chunk_buffer:, task_logger:, dispatch_interval:, logger:)
  @event_queue = event_queue
  @task_queue = task_queue
  @chunk_router = chunk_router
  @chunk_buffer = chunk_buffer
  @task_logger = task_logger
  @dispatch_interval = dispatch_interval
  @dispatch_message_id = nil
  @logger = logger
  @dispatch_requested = false
  @checkpoint_requested = false
end

Instance Attribute Details

#logger ⇒ Object (readonly)

Returns the value of attribute logger.



118
119
120
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 118

def logger
  @logger
end

Class Method Details

._main ⇒ Object 



29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 29

def Dispatcher._main
  opts = DispatcherOptions.new(ARGV)
  opts.parse
  unless opts.rest_arguments.size == 1
    $stderr.puts opts.usage
    exit 1
  end
  config_path, * = opts.rest_arguments
  config = YAML.load(File.read(config_path))
  log = opts.log_file_path ? new_logger(File.expand_path(opts.log_file_path), config) : nil
  ctx = Context.for_application('.', environment: opts.environment, logger: log)
  logger = raw_logger = ctx.logger
  event_queue = ctx.get_data_source('sqs', config.fetch('event-queue-ds', 'sqs_event'))
  task_queue = ctx.get_data_source('sqs', config.fetch('task-queue-ds', 'sqs_task'))
  if config['alert-level']
    logger = AlertingLogger.new(
      logger: raw_logger,
      sns_datasource: ctx.get_data_source('sns', config.fetch('sns-ds', 'sns')),
      alert_level: config.fetch('alert-level', 'warn')
    )
  end

  chunk_buffer = ChunkBuffer.new(
    control_data_source: ctx.get_data_source('sql', config.fetch('ctl-postgres-ds', 'db_ctl')),
    logger: logger
  )

  chunk_router = ChunkRouter.for_config(config.fetch('url_patterns'))

  task_logger = LoadTaskLogger.new(
    ctx.get_data_source('s3', config.fetch('ctl-s3-ds', 's3_ctl'))
  )

  dispatcher = Dispatcher.new(
    event_queue: event_queue,
    task_queue: task_queue,
    chunk_router: chunk_router,
    chunk_buffer: chunk_buffer,
    task_logger: task_logger,
    dispatch_interval: config.fetch('dispatch-interval', 60),
    logger: logger
  )

  if opts.task_id
    dispatcher.dispatch_tasks chunk_buffer.load_tasks_by_id([opts.task_id])
    exit 0
  end

  Process.daemon(true) if opts.daemon?
  create_pid_file opts.pid_file_path if opts.pid_file_path
  Dir.chdir '/'
  begin
    dispatcher.event_loop
  rescue Exception => e
    logger.exception e
    logger.error "dispatcher abort: pid=#{$$}"
    raise
  end
end

.create_pid_file(path) ⇒ Object 



97
98
99
100
101
102
103
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 97

def Dispatcher.create_pid_file(path)
  File.open(path, 'w') {|f|
    f.puts $$
  }
rescue
  # ignore
end

.main ⇒ Object 



23
24
25
26
27
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 23

def Dispatcher.main
  Raven.capture {
    _main
  }
end

.new_logger(path, config) ⇒ Object 



89
90
91
92
93
94
95
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 89

def Dispatcher.new_logger(path, config)
  Logger.new(
    device: path,
    rotation_period: config.fetch('log-rotation-period', 'daily'),
    rotation_size: config.fetch('log-rotation-size', nil)
  )
end

Instance Method Details

#after_message_batch ⇒ Object

override



129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 129

def after_message_batch
  # must be processed first
  @event_queue.process_async_delete

  if @dispatch_requested
    logger.info "*** dispatch requested"
    do_handle_dispatch
    @dispatch_requested = false
  end

  if @checkpoint_requested
    do_handle_checkpoint
    @checkpoint_requested = false   # is needless, but reset it just in case
  end
end

#dispatch_tasks(tasks) ⇒ Object 



215
216
217
218
219
220
221
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 215

def dispatch_tasks(tasks)
  tasks.each do |task|
    msg = StreamingLoadV3LoaderMessage.for_load_task(task)
    @task_queue.put msg
    @task_logger.log task
  end
end

#do_handle_checkpoint ⇒ Object 



165
166
167
168
169
170
171
172
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 165

def do_handle_checkpoint
  logger.info "*** checkpoint requested"
  logger.info "Force-flushing all objects..."
  tasks = @chunk_buffer.flush_all
  dispatch_tasks tasks
  logger.info "All objects flushed; shutting down..."
  @event_queue.initiate_terminate
end

#do_handle_dispatch ⇒ Object 



193
194
195
196
197
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 193

def do_handle_dispatch
  tasks = @chunk_buffer.flush_partial
  dispatch_tasks tasks
  set_dispatch_timer
end

#event_loop ⇒ Object 



120
121
122
123
124
125
126
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 120

def event_loop
  logger.info "*** dispatcher started: pid=#{$$}"
  set_dispatch_timer
  @event_queue.handle_messages(handler: self, message_class: DispatcherMessage)
  @event_queue.process_async_delete_force
  logger.info "*** shutdown gracefully: pid=#{$$}"
end

#handle_checkpoint(msg) ⇒ Object 



157
158
159
160
161
162
163
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 157

def handle_checkpoint(msg)
  # Delay creating CHECKPOINT after the current message batch,
  # because any other extra events are already received.
  @checkpoint_requested = true
  # Delete this event immediately
  @event_queue.delete_message(msg)
end

#handle_data(msg) ⇒ Object 



174
175
176
177
178
179
180
181
182
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 174

def handle_data(msg)
  unless msg.created_event?
    @event_queue.delete_message_async(msg)
    return
  end
  chunk = @chunk_router.route(msg)
  @chunk_buffer.save(chunk)
  @event_queue.delete_message_async(msg)
end

#handle_dispatch(msg) ⇒ Object 



184
185
186
187
188
189
190
191
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 184

def handle_dispatch(msg)
  # Dispatching tasks may takes 10 minutes or more, it can exceeds visibility timeout.
  # To avoid this, delay dispatching until all events of current message batch are processed.
  if @dispatch_message_id == msg.message_id
    @dispatch_requested = true
  end
  @event_queue.delete_message_async(msg)
end

#handle_flushtable(msg) ⇒ Object 



204
205
206
207
208
209
210
211
212
213
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 204

def handle_flushtable(msg)
  # FIXME: badly named attribute. table_name is really stream_name, which is called as data_source_id, too.
  stream_name = msg.table_name

  logger.info "*** flushtable requested: stream_name=#{stream_name}"
  tasks = @chunk_buffer.flush_stream(stream_name)
  dispatch_tasks tasks
  # Delete this event immediately
  @event_queue.delete_message(msg)
end

#handle_shutdown(msg) ⇒ Object 



150
151
152
153
154
155
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 150

def handle_shutdown(msg)
  logger.info "*** shutdown requested"
  @event_queue.initiate_terminate
  # Delete this event immediately
  @event_queue.delete_message(msg)
end

#handle_unknown(msg) ⇒ Object 



145
146
147
148
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 145

def handle_unknown(msg)
  logger.warn "unknown event: #{msg.message_body}"
  @event_queue.delete_message_async(msg)
end

#set_dispatch_timer ⇒ Object 



199
200
201
202
 
# File 'lib/bricolage/streamingload/dispatcher.rb', line 199

def set_dispatch_timer
  res = @event_queue.send_message(DispatchDispatcherMessage.create(delay_seconds: @dispatch_interval))
  @dispatch_message_id = res.message_id
end