ruby-kafka 1.0.0 → 1.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8ebe10cb2c93b4387a14d069c4291751808abed924cbee6f4ff2f00770aaa1c2
-  data.tar.gz: 1e392b4f2b5137cd83bd191b5c629fd98b41697d389fb5c68c3100d03c3ff218
+  metadata.gz: 22b958f3d452f35c3c6084a5b8b790f370b96069e87ef5974838e5ea4b6945e1
+  data.tar.gz: eb6d704dbaace9c13a99bc93924d8a6b84ee5eb686c9c863a81d085ffeb7e92d
 SHA512:
-  metadata.gz: 037a8fe1a7495f5e3d723e047392f68a1373e91f8a00337beb7ae6e053d031d6ab1b8597767a1b8b43399850e23493295180a24250e0c2f09b4f626316cd6915
-  data.tar.gz: 108d7160593e452f27c1e828438f4cd8c2c6c009f081a0ff949ca330f9487b948aad73cf23ba39e12adb0ed9f773c667cfeebaad1625b33770057b7b69b3dcb2
+  metadata.gz: 22c1d59bcdd42849849122b559f0d161653a5cfa492ffb36f28a875b109444d670a76bc62ba0856ae2ebce95fd10abce9e16834928fc7fe6eb2ee006827d307c
+  data.tar.gz: 3a51df6b1d40e1edbd96c06f3473319f0ac1b072b040b676a8e1d980fbaad114ad0248a7670cb7ae0a51a2c79ac390f2f4cfe8f65eedb5dd55be020c6011bc18

data/.circleci/config.yml

@@ -267,6 +267,38 @@ jobs:
       - run: bundle install --path vendor/bundle
       - run: bundle exec rspec --profile --tag functional spec/functional
 
+  kafka-2.5:
+    docker:
+      - image: circleci/ruby:2.5.1-node
+        environment:
+          LOG_LEVEL: DEBUG
+      - image: wurstmeister/zookeeper
+      - image: wurstmeister/kafka:2.12-2.5.0
+        environment:
+          KAFKA_ADVERTISED_HOST_NAME: localhost
+          KAFKA_ADVERTISED_PORT: 9092
+          KAFKA_PORT: 9092
+          KAFKA_ZOOKEEPER_CONNECT: localhost:2181
+          KAFKA_DELETE_TOPIC_ENABLE: true
+      - image: wurstmeister/kafka:2.12-2.5.0
+        environment:
+          KAFKA_ADVERTISED_HOST_NAME: localhost
+          KAFKA_ADVERTISED_PORT: 9093
+          KAFKA_PORT: 9093
+          KAFKA_ZOOKEEPER_CONNECT: localhost:2181
+          KAFKA_DELETE_TOPIC_ENABLE: true
+      - image: wurstmeister/kafka:2.12-2.5.0
+        environment:
+          KAFKA_ADVERTISED_HOST_NAME: localhost
+          KAFKA_ADVERTISED_PORT: 9094
+          KAFKA_PORT: 9094
+          KAFKA_ZOOKEEPER_CONNECT: localhost:2181
+          KAFKA_DELETE_TOPIC_ENABLE: true
+    steps:
+      - checkout
+      - run: bundle install --path vendor/bundle
+      - run: bundle exec rspec --profile --tag functional spec/functional
+
 workflows:
   version: 2
   test:
@@ -280,3 +312,4 @@ workflows:
       - kafka-2.2
       - kafka-2.3
       - kafka-2.4
+      - kafka-2.5

data/CHANGELOG.md

@@ -4,6 +4,16 @@ Changes and additions to the library will be listed here.
 
 ## Unreleased
 
+## 1.1.0
+
+- Extra sanity checking when marking offsets as processed (#824).
+- Make `verify_hostname` settable for SSL contexts (#828).
+- Instrument `create_time` from last message in batch (#811).
+- Add client function for fetching topic replica count (#822).
+- Allow consumers to refresh the topic lists (#818).
+- Disconnect after leaving a group (#817).
+- Use `max_wait_time` as the sleep instead of hardcoded 2 seconds (#825).
+
 ## 1.0.0
 
 - Add client methods to manage configs (#759)

data/README.md

@@ -118,10 +118,16 @@ Or install it yourself as:
     <td>Limited support</td>
     <td>Limited support</td>
   </tr>
+  <tr>
     <th>Kafka 2.4</th>
     <td>Limited support</td>
     <td>Limited support</td>
   </tr>
+  <tr>
+    <th>Kafka 2.5</th>
+    <td>Limited support</td>
+    <td>Limited support</td>
+  </tr>
 </table>
 
 This library is targeting Kafka 0.9 with the v0.4.x series and Kafka 0.10 with the v0.5.x series. There's limited support for Kafka 0.8, and things should work with Kafka 0.11, although there may be performance issues due to changes in the protocol.
@@ -136,6 +142,7 @@ This library is targeting Kafka 0.9 with the v0.4.x series and Kafka 0.10 with t
 - **Kafka 2.2:** Everything that works with Kafka 2.1 should still work, but so far no features specific to Kafka 2.2 have been added.
 - **Kafka 2.3:** Everything that works with Kafka 2.2 should still work, but so far no features specific to Kafka 2.3 have been added.
 - **Kafka 2.4:** Everything that works with Kafka 2.3 should still work, but so far no features specific to Kafka 2.4 have been added.
+- **Kafka 2.5:** Everything that works with Kafka 2.4 should still work, but so far no features specific to Kafka 2.5 have been added.
 
 This library requires Ruby 2.1 or higher.
 
@@ -945,6 +952,8 @@ This configures the store to look up CA certificates from the system default cer
 
 In order to authenticate the client to the cluster, you need to pass in a certificate and key created for the client and trusted by the brokers.
 
+**NOTE**: You can disable hostname validation by passing `verify_hostname: false`.
+
 ```ruby
 kafka = Kafka.new(
   ["kafka1:9092"],

data/lib/kafka/client.rb

@@ -65,6 +65,10 @@ module Kafka
     # @param sasl_oauth_token_provider [Object, nil] OAuthBearer Token Provider instance that
     #   implements method token. See {Sasl::OAuth#initialize}
     #
+    # @param verify_hostname [Boolean, true] whether to verify that the host serving
+    #   the SSL certificate and the signing chain of the certificate have the correct domains
+    #   based on the CA certificate
+    #
     # @return [Client]
     def initialize(seed_brokers:, client_id: "ruby-kafka", logger: nil, connect_timeout: nil, socket_timeout: nil,
                    ssl_ca_cert_file_path: nil, ssl_ca_cert: nil, ssl_client_cert: nil, ssl_client_cert_key: nil,
@@ -336,6 +340,9 @@ module Kafka
     # @param fetcher_max_queue_size [Integer] max number of items in the fetch queue that
     #   are stored for further processing. Note, that each item in the queue represents a
     #   response from a single broker.
+    # @param refresh_topic_interval [Integer] interval of refreshing the topic list.
+    #   If it is 0, the topic list won't be refreshed (default)
+    #   If it is n (n > 0), the topic list will be refreshed every n seconds
     # @return [Consumer]
     def consumer(
         group_id:,
@@ -345,7 +352,8 @@ module Kafka
         offset_commit_threshold: 0,
         heartbeat_interval: 10,
         offset_retention_time: nil,
-        fetcher_max_queue_size: 100
+        fetcher_max_queue_size: 100,
+        refresh_topic_interval: 0
     )
       cluster = initialize_cluster
 
@@ -399,6 +407,7 @@ module Kafka
         fetcher: fetcher,
         session_timeout: session_timeout,
         heartbeat: heartbeat,
+        refresh_topic_interval: refresh_topic_interval
       )
     end
 
@@ -694,6 +703,14 @@ module Kafka
       @cluster.partitions_for(topic).count
     end
 
+    # Counts the number of replicas for a topic's partition
+    #
+    # @param topic [String]
+    # @return [Integer] the number of replica nodes for the topic's partition
+    def replica_count_for(topic)
+      @cluster.partitions_for(topic).first.replicas.count
+    end
+
     # Retrieve the offset of the last message in a partition. If there are no
     # messages in the partition -1 is returned.
     #

data/lib/kafka/consumer.rb

@@ -44,7 +44,7 @@ module Kafka
   #
   class Consumer
 
-    def initialize(cluster:, logger:, instrumenter:, group:, fetcher:, offset_manager:, session_timeout:, heartbeat:)
+    def initialize(cluster:, logger:, instrumenter:, group:, fetcher:, offset_manager:, session_timeout:, heartbeat:, refresh_topic_interval: 0)
       @cluster = cluster
       @logger = TaggedLogger.new(logger)
       @instrumenter = instrumenter
@@ -53,6 +53,7 @@ module Kafka
       @session_timeout = session_timeout
       @fetcher = fetcher
       @heartbeat = heartbeat
+      @refresh_topic_interval = refresh_topic_interval
 
       @pauses = Hash.new {|h, k|
         h[k] = Hash.new {|h2, k2|
@@ -73,6 +74,15 @@ module Kafka
       #   when user commits message other than last in a batch, this would make ruby-kafka refetch
       #   some already consumed messages
       @current_offsets = Hash.new { |h, k| h[k] = {} }
+
+      # Map storing subscribed topics with their configuration
+      @subscribed_topics = Concurrent::Map.new
+
+      # Set storing topics that matched topics in @subscribed_topics
+      @matched_topics = Set.new
+
+      # Whether join_group must be executed again because new topics are added
+      @join_group_for_new_topics = false
     end
 
     # Subscribes the consumer to a topic.
@@ -97,13 +107,12 @@ module Kafka
     def subscribe(topic_or_regex, default_offset: nil, start_from_beginning: true, max_bytes_per_partition: 1048576)
       default_offset ||= start_from_beginning ? :earliest : :latest
 
-      if topic_or_regex.is_a?(Regexp)
-        cluster_topics.select { |topic| topic =~ topic_or_regex }.each do |topic|
-          subscribe_to_topic(topic, default_offset, start_from_beginning, max_bytes_per_partition)
-        end
-      else
-        subscribe_to_topic(topic_or_regex, default_offset, start_from_beginning, max_bytes_per_partition)
-      end
+      @subscribed_topics[topic_or_regex] = {
+        default_offset: default_offset,
+        start_from_beginning: start_from_beginning,
+        max_bytes_per_partition: max_bytes_per_partition
+      }
+      scan_for_subscribing
 
       nil
     end
@@ -116,7 +125,6 @@ module Kafka
     def stop
       @running = false
       @fetcher.stop
-      @cluster.disconnect
     end
 
     # Pause processing of a specific topic partition.
@@ -308,6 +316,7 @@ module Kafka
               topic: batch.topic,
               partition: batch.partition,
               last_offset: batch.last_offset,
+              last_create_time: batch.messages.last.try(:create_time),
               offset_lag: batch.offset_lag,
               highwater_mark_offset: batch.highwater_mark_offset,
               message_count: batch.messages.count,
@@ -401,6 +410,7 @@ module Kafka
       while running?
         begin
           @instrumenter.instrument("loop.consumer") do
+            refresh_topic_list_if_enabled
             yield
           end
         rescue HeartbeatError
@@ -432,6 +442,7 @@ module Kafka
       # important that members explicitly tell Kafka when they're leaving.
       make_final_offsets_commit!
       @group.leave rescue nil
+      @cluster.disconnect
       @running = false
       @logger.pop_tags
     end
@@ -452,6 +463,8 @@ module Kafka
     end
 
     def join_group
+      @join_group_for_new_topics = false
+
       old_generation_id = @group.generation_id
 
       @group.join
@@ -513,11 +526,19 @@ module Kafka
       end
     end
 
+    def refresh_topic_list_if_enabled
+      return if @refresh_topic_interval <= 0
+      return if @refreshed_at && @refreshed_at + @refresh_topic_interval > Time.now
+
+      scan_for_subscribing
+      @refreshed_at = Time.now
+    end
+
     def fetch_batches
       # Return early if the consumer has been stopped.
       return [] if shutting_down?
 
-      join_group unless @group.member?
+      join_group if !@group.member? || @join_group_for_new_topics
 
       trigger_heartbeat
 
@@ -525,7 +546,7 @@ module Kafka
 
       if !@fetcher.data?
         @logger.debug "No batches to process"
-        sleep 2
+        sleep(@fetcher.max_wait_time || 2)
         []
       else
         tag, message = @fetcher.poll
@@ -571,10 +592,34 @@ module Kafka
       end
     end
 
+    def scan_for_subscribing
+      @subscribed_topics.each do |topic_or_regex, config|
+        default_offset = config.fetch(:default_offset)
+        start_from_beginning = config.fetch(:start_from_beginning)
+        max_bytes_per_partition = config.fetch(:max_bytes_per_partition)
+        if topic_or_regex.is_a?(Regexp)
+          subscribe_to_regex(topic_or_regex, default_offset, start_from_beginning, max_bytes_per_partition)
+        else
+          subscribe_to_topic(topic_or_regex, default_offset, start_from_beginning, max_bytes_per_partition)
+        end
+      end
+    end
+
+    def subscribe_to_regex(topic_regex, default_offset, start_from_beginning, max_bytes_per_partition)
+      cluster_topics.select { |topic| topic =~ topic_regex }.each do |topic|
+        subscribe_to_topic(topic, default_offset, start_from_beginning, max_bytes_per_partition)
+      end
+    end
+
     def subscribe_to_topic(topic, default_offset, start_from_beginning, max_bytes_per_partition)
+      return if @matched_topics.include?(topic)
+      @matched_topics.add(topic)
+      @join_group_for_new_topics = true
+
       @group.subscribe(topic)
       @offset_manager.set_default_offset(topic, default_offset)
       @fetcher.subscribe(topic, max_bytes_per_partition: max_bytes_per_partition)
+      @cluster.mark_as_stale!
     end
 
     def cluster_topics

data/lib/kafka/consumer_group.rb

@@ -161,6 +161,12 @@ module Kafka
       @member_id = ""
       sleep 1
 
+      retry
+    rescue CoordinatorLoadInProgress
+      @logger.error "Coordinator broker still loading, retrying in 1s..."
+
+      sleep 1
+
       retry
     end
 

data/lib/kafka/datadog.rb

@@ -31,7 +31,7 @@ module Kafka
 
     class << self
       def statsd
-        @statsd ||= ::Datadog::Statsd.new(host, port, namespace: namespace, tags: tags)
+        @statsd ||= ::Datadog::Statsd.new(host, port, namespace: namespace, tags: tags, socket_path: socket_path)
       end
 
       def statsd=(statsd)
@@ -57,6 +57,15 @@ module Kafka
         clear
       end
 
+      def socket_path
+        @socket_path
+      end
+
+      def socket_path=(socket_path)
+        @socket_path = socket_path
+        clear
+      end
+
       def namespace
         @namespace ||= STATSD_NAMESPACE
       end
@@ -160,6 +169,8 @@ module Kafka
       def process_batch(event)
         offset = event.payload.fetch(:last_offset)
         messages = event.payload.fetch(:message_count)
+        create_time = event.payload.fetch(:last_create_time)
+        time_lag = create_time && ((Time.now - create_time) * 1000).to_i
 
         tags = {
           client: event.payload.fetch(:client_id),
@@ -176,6 +187,10 @@ module Kafka
         end
 
         gauge("consumer.offset", offset, tags: tags)
+
+        if time_lag
+          gauge("consumer.time_lag", time_lag, tags: tags)
+        end
       end
 
       def fetch_batch(event)

data/lib/kafka/fetcher.rb

@@ -4,7 +4,7 @@ require "kafka/fetch_operation"
 
 module Kafka
   class Fetcher
-    attr_reader :queue
+    attr_reader :queue, :max_wait_time
 
     def initialize(cluster:, logger:, instrumenter:, max_queue_size:, group:)
       @cluster = cluster

data/lib/kafka/offset_manager.rb

@@ -50,9 +50,20 @@ module Kafka
     # @param offset [Integer] the offset of the message that should be marked as processed.
     # @return [nil]
     def mark_as_processed(topic, partition, offset)
-      @uncommitted_offsets += 1
+      unless @group.assigned_to?(topic, partition)
+        @logger.debug "Not marking #{topic}/#{partition}:#{offset} as processed for partition not assigned to this consumer."
+        return
+      end
       @processed_offsets[topic] ||= {}
 
+      last_processed_offset = @processed_offsets[topic][partition] || -1
+      if last_processed_offset > offset + 1
+        @logger.debug "Not overwriting newer offset #{topic}/#{partition}:#{last_processed_offset - 1} with older #{offset}"
+        return
+      end
+
+      @uncommitted_offsets += 1
+
       # The committed offset should always be the offset of the next message that the
       # application will read, thus adding one to the last message processed.
       @processed_offsets[topic][partition] = offset + 1

data/lib/kafka/protocol/metadata_response.rb

@@ -34,7 +34,7 @@ module Kafka
     #
     class MetadataResponse
       class PartitionMetadata
-        attr_reader :partition_id, :leader
+        attr_reader :partition_id, :leader, :replicas
 
         attr_reader :partition_error_code
 

data/lib/kafka/ssl_context.rb

@@ -54,11 +54,12 @@ module Kafka
           store.set_default_paths
         end
         ssl_context.cert_store = store
-        ssl_context.verify_mode = OpenSSL::SSL::VERIFY_PEER
-        # Verify certificate hostname if supported (ruby >= 2.4.0)
-        ssl_context.verify_hostname = verify_hostname if ssl_context.respond_to?(:verify_hostname=)
       end
 
+      ssl_context.verify_mode = OpenSSL::SSL::VERIFY_PEER
+      # Verify certificate hostname if supported (ruby >= 2.4.0)
+      ssl_context.verify_hostname = verify_hostname if ssl_context.respond_to?(:verify_hostname=)
+
       ssl_context
     end
   end

data/lib/kafka/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kafka
-  VERSION = "1.0.0"
+  VERSION = "1.1.0"
 end

data/ruby-kafka.gemspec

@@ -18,7 +18,7 @@ Gem::Specification.new do |spec|
   DESC
 
   spec.homepage      = "https://github.com/zendesk/ruby-kafka"
-  spec.license       = "Apache License Version 2.0"
+  spec.license       = "Apache-2.0"
 
   spec.required_ruby_version = '>= 2.1.0'
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ruby-kafka
 version: !ruby/object:Gem::Version
-  version: 1.0.0
+  version: 1.1.0
 platform: ruby
 authors:
 - Daniel Schierbeck
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2020-02-25 00:00:00.000000000 Z
+date: 2020-06-09 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: digest-crc
@@ -476,7 +476,7 @@ files:
 - ruby-kafka.gemspec
 homepage: https://github.com/zendesk/ruby-kafka
 licenses:
-- Apache License Version 2.0
+- Apache-2.0
 metadata: {}
 post_install_message: 
 rdoc_options: []