Class: Kafka::Cluster

Inherits:
Object
  • Object
show all
Defined in:
lib/kafka/cluster.rb

Overview

A cluster represents the state of a Kafka cluster. It needs to be initialized with a non-empty list of seed brokers. The first seed broker that the cluster can connect to will be asked for the cluster metadata, which allows the cluster to map topic partitions to the current leader for those partitions.

Instance Method Summary collapse

Constructor Details

#initialize(seed_brokers:, broker_pool:, logger:) ⇒ Cluster

Initializes a Cluster with a set of seed brokers.

The cluster will try to fetch cluster metadata from one of the brokers.

Parameters:


21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
# File 'lib/kafka/cluster.rb', line 21

def initialize(seed_brokers:, broker_pool:, logger:)
  if seed_brokers.empty?
    raise ArgumentError, "At least one seed broker must be configured"
  end

  @logger = TaggedLogger.new(logger)
  @seed_brokers = seed_brokers
  @broker_pool = broker_pool
  @cluster_info = nil
  @stale = true

  # This is the set of topics we need metadata for. If empty, metadata for
  # all topics will be fetched.
  @target_topics = Set.new
end

Instance Method Details

#add_target_topics(topics) ⇒ nil

Adds a list of topics to the target list. Only the topics on this list will be queried for metadata.

Parameters:

  • topics (Array<String>)

Returns:

  • (nil)

42
43
44
45
46
47
48
49
50
51
52
53
54
55
# File 'lib/kafka/cluster.rb', line 42

def add_target_topics(topics)
  topics = Set.new(topics)
  unless topics.subset?(@target_topics)
    new_topics = topics - @target_topics

    unless new_topics.empty?
      @logger.info "New topics added to target list: #{new_topics.to_a.join(', ')}"

      @target_topics.merge(new_topics)

      refresh_metadata!
    end
  end
end

#alter_topic(name, configs = {}) ⇒ Object


230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
# File 'lib/kafka/cluster.rb', line 230

def alter_topic(name, configs = {})
  options = {
    resources: [[Kafka::Protocol::RESOURCE_TYPE_TOPIC, name, configs]]
  }

  broker = controller_broker

  @logger.info "Altering the config for topic `#{name}` using controller broker #{broker}"

  response = broker.alter_configs(**options)

  response.resources.each do |resource|
    Protocol.handle_error(resource.error_code, resource.error_message)
  end

  nil
end

#api_info(api_key) ⇒ Object


57
58
59
# File 'lib/kafka/cluster.rb', line 57

def api_info(api_key)
  apis.find {|api| api.api_key == api_key }
end

#apisObject


72
73
74
75
76
77
78
79
80
81
# File 'lib/kafka/cluster.rb', line 72

def apis
  @apis ||=
    begin
      response = random_broker.api_versions

      Protocol.handle_error(response.error_code)

      response.apis
    end
end

#clear_target_topicsnil

Clears the list of target topics.

Returns:

  • (nil)

See Also:


87
88
89
90
# File 'lib/kafka/cluster.rb', line 87

def clear_target_topics
  @target_topics.clear
  refresh_metadata!
end

#cluster_infoObject


349
350
351
# File 'lib/kafka/cluster.rb', line 349

def cluster_info
  @cluster_info ||= fetch_cluster_info
end

#create_partitions_for(name, num_partitions:, timeout:) ⇒ Object


255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
# File 'lib/kafka/cluster.rb', line 255

def create_partitions_for(name, num_partitions:, timeout:)
  options = {
    topics: [[name, num_partitions, nil]],
    timeout: timeout
  }

  broker = controller_broker

  @logger.info "Creating #{num_partitions} partition(s) for topic `#{name}` using controller broker #{broker}"

  response = broker.create_partitions(**options)

  response.errors.each do |topic, error_code, error_message|
    Protocol.handle_error(error_code, error_message)
  end
  mark_as_stale!

  @logger.info "Topic `#{name}` was updated"
end

#create_topic(name, num_partitions:, replication_factor:, timeout:, config:) ⇒ Object


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
177
178
179
180
181
182
183
184
185
186
187
188
189
190
# File 'lib/kafka/cluster.rb', line 151

def create_topic(name, num_partitions:, replication_factor:, timeout:, config:)
  options = {
    topics: {
      name => {
        num_partitions: num_partitions,
        replication_factor: replication_factor,
        config: config,
      }
    },
    timeout: timeout,
  }

  broker = controller_broker

  @logger.info "Creating topic `#{name}` using controller broker #{broker}"

  response = broker.create_topics(**options)

  response.errors.each do |topic, error_code|
    Protocol.handle_error(error_code)
  end

  begin
    partitions_for(name).each do |info|
      Protocol.handle_error(info.partition_error_code)
    end
  rescue Kafka::LeaderNotAvailable
    @logger.warn "Leader not yet available for `#{name}`, waiting 1s..."
    sleep 1

    retry
  rescue Kafka::UnknownTopicOrPartition
    @logger.warn "Topic `#{name}` not yet created, waiting 1s..."
    sleep 1

    retry
  end

  @logger.info "Topic `#{name}` was created"
end

#delete_topic(name, timeout:) ⇒ Object


192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
# File 'lib/kafka/cluster.rb', line 192

def delete_topic(name, timeout:)
  options = {
    topics: [name],
    timeout: timeout,
  }

  broker = controller_broker

  @logger.info "Deleting topic `#{name}` using controller broker #{broker}"

  response = broker.delete_topics(**options)

  response.errors.each do |topic, error_code|
    Protocol.handle_error(error_code)
  end

  @logger.info "Topic `#{name}` was deleted"
end

#describe_group(group_id) ⇒ Object


248
249
250
251
252
253
# File 'lib/kafka/cluster.rb', line 248

def describe_group(group_id)
  response = get_group_coordinator(group_id: group_id).describe_groups(group_ids: [group_id])
  group = response.groups.first
  Protocol.handle_error(group.error_code)
  group
end

#describe_topic(name, configs = []) ⇒ Object


211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
# File 'lib/kafka/cluster.rb', line 211

def describe_topic(name, configs = [])
  options = {
    resources: [[Kafka::Protocol::RESOURCE_TYPE_TOPIC, name, configs]]
  }
  broker = controller_broker

  @logger.info "Fetching topic `#{name}`'s configs using controller broker #{broker}"

  response = broker.describe_configs(**options)

  response.resources.each do |resource|
    Protocol.handle_error(resource.error_code, resource.error_message)
  end
  topic_description = response.resources.first
  topic_description.configs.each_with_object({}) do |config, hash|
    hash[config.name] = config.value
  end
end

#disconnectObject


345
346
347
# File 'lib/kafka/cluster.rb', line 345

def disconnect
  @broker_pool.close
end

#get_group_coordinator(group_id:) ⇒ Broker

Finds the broker acting as the coordinator of the given group.

Parameters:

  • group_id: (String)

Returns:

  • (Broker)

    the broker that's currently coordinator.


118
119
120
121
122
# File 'lib/kafka/cluster.rb', line 118

def get_group_coordinator(group_id:)
  @logger.debug "Getting group coordinator for `#{group_id}`"
  
  get_coordinator(Kafka::Protocol::COORDINATOR_TYPE_GROUP, group_id)
end

#get_leader(topic, partition) ⇒ Broker

Finds the broker acting as the leader of the given topic and partition.

Parameters:

  • topic (String)
  • partition (Integer)

Returns:

  • (Broker)

    the broker that's currently leader.


110
111
112
# File 'lib/kafka/cluster.rb', line 110

def get_leader(topic, partition)
  connect_to_broker(get_leader_id(topic, partition))
end

#get_transaction_coordinator(transactional_id:) ⇒ Broker

Finds the broker acting as the coordinator of the given transaction.

Parameters:

  • transactional_id: (String)

Returns:

  • (Broker)

    the broker that's currently coordinator.


128
129
130
131
132
133
134
135
136
137
138
139
140
# File 'lib/kafka/cluster.rb', line 128

def get_transaction_coordinator(transactional_id:)
  @logger.debug "Getting transaction coordinator for `#{transactional_id}`"

  

  if transactional_id.nil?
    # Get a random_broker
    @logger.debug "Transaction ID is not available. Choose a random broker."
    return random_broker
  else
    get_coordinator(Kafka::Protocol::COORDINATOR_TYPE_TRANSACTION, transactional_id)
  end
end

#list_groupsObject


336
337
338
339
340
341
342
343
# File 'lib/kafka/cluster.rb', line 336

def list_groups
  
  cluster_info.brokers.map do |broker|
    response = connect_to_broker(broker.node_id).list_groups
    Protocol.handle_error(response.error_code)
    response.groups.map(&:group_id)
  end.flatten.uniq
end

#list_topicsObject

Lists all topics in the cluster.


329
330
331
332
333
334
# File 'lib/kafka/cluster.rb', line 329

def list_topics
  response = random_broker.(topics: nil)
  response.topics.select do |topic|
    topic.topic_error_code == 0
  end.map(&:topic_name)
end

#mark_as_stale!Object


92
93
94
# File 'lib/kafka/cluster.rb', line 92

def mark_as_stale!
  @stale = true
end

#partitions_for(topic) ⇒ Object


142
143
144
145
146
147
148
149
# File 'lib/kafka/cluster.rb', line 142

def partitions_for(topic)
  add_target_topics([topic])
  
  cluster_info.partitions_for(topic)
rescue Kafka::ProtocolError
  mark_as_stale!
  raise
end

#refresh_metadata!Object


96
97
98
99
# File 'lib/kafka/cluster.rb', line 96

def refresh_metadata!
  @cluster_info = nil
  cluster_info
end

#refresh_metadata_if_necessary!Object


101
102
103
# File 'lib/kafka/cluster.rb', line 101

def 
  refresh_metadata! if @stale
end

#resolve_offset(topic, partition, offset) ⇒ Object


317
318
319
# File 'lib/kafka/cluster.rb', line 317

def resolve_offset(topic, partition, offset)
  resolve_offsets(topic, [partition], offset).fetch(partition)
end

#resolve_offsets(topic, partitions, offset) ⇒ Object


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
# File 'lib/kafka/cluster.rb', line 275

def resolve_offsets(topic, partitions, offset)
  add_target_topics([topic])
  

  partitions_by_broker = partitions.each_with_object({}) {|partition, hsh|
    broker = get_leader(topic, partition)

    hsh[broker] ||= []
    hsh[broker] << partition
  }

  if offset == :earliest
    offset = -2
  elsif offset == :latest
    offset = -1
  end

  offsets = {}

  partitions_by_broker.each do |broker, broker_partitions|
    response = broker.list_offsets(
      topics: {
        topic => broker_partitions.map {|partition|
          {
            partition: partition,
            time: offset
          }
        }
      }
    )

    broker_partitions.each do |partition|
      offsets[partition] = response.offset_for(topic, partition)
    end
  end

  offsets
rescue Kafka::ProtocolError
  mark_as_stale!
  raise
end

#supports_api?(api_key, version = nil) ⇒ Boolean

Returns:

  • (Boolean)

61
62
63
64
65
66
67
68
69
70
# File 'lib/kafka/cluster.rb', line 61

def supports_api?(api_key, version = nil)
  info = api_info(api_key)
  if info.nil?
    return false
  elsif version.nil?
    return true
  else
    return info.version_supported?(version)
  end
end

#topicsObject


321
322
323
324
325
326
# File 'lib/kafka/cluster.rb', line 321

def topics
  
  cluster_info.topics.select do |topic|
    topic.topic_error_code == 0
  end.map(&:topic_name)
end