ruby-kafka 1.0.0 → 1.1.0.beta1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8ebe10cb2c93b4387a14d069c4291751808abed924cbee6f4ff2f00770aaa1c2
-  data.tar.gz: 1e392b4f2b5137cd83bd191b5c629fd98b41697d389fb5c68c3100d03c3ff218
+  metadata.gz: 1e5516c0741a1532bd320d8202e4cd23d18af5ddaf9581c6a39c69230b4c59e1
+  data.tar.gz: 66b47c941718a9a3a6e9f0523ec60dc6dd8b96ef2c82fda3571f7d150464cfeb
 SHA512:
-  metadata.gz: 037a8fe1a7495f5e3d723e047392f68a1373e91f8a00337beb7ae6e053d031d6ab1b8597767a1b8b43399850e23493295180a24250e0c2f09b4f626316cd6915
-  data.tar.gz: 108d7160593e452f27c1e828438f4cd8c2c6c009f081a0ff949ca330f9487b948aad73cf23ba39e12adb0ed9f773c667cfeebaad1625b33770057b7b69b3dcb2
+  metadata.gz: cb040bc7a76c13f131b59e2a15d9274542a4768595de221198b4e574aeb36ff5095ac14b0737b6a988cd4be7863f4bd44cccc7d2ad48ae2ee3f72f81256b08ca
+  data.tar.gz: 012474c8b6390eb2348906f320caa3759c12d8e9205713cd173f98a1463a432caa8c1c5cd700a2bea059e6601c5482b158d72af1fb199d3f9751f3cb1a3b5451

data/CHANGELOG.md

@@ -4,6 +4,14 @@ Changes and additions to the library will be listed here.
 
 ## Unreleased
 
+## 1.1.0
+
+- 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/lib/kafka/client.rb

@@ -336,6 +336,9 @@ 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 +348,8 @@ 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 +403,7 @@ module Kafka
         fetcher: fetcher,
         session_timeout: session_timeout,
         heartbeat: heartbeat,
+        refresh_topic_interval: refresh_topic_interval
       )
     end
 
@@ -694,6 +699,14 @@ 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/datadog.rb

@@ -160,6 +160,8 @@ 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 +178,10 @@ 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/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/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kafka
-  VERSION = "1.0.0"
+  VERSION = "1.1.0.beta1"
 end

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.beta1
 platform: ruby
 authors:
 - Daniel Schierbeck
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2020-02-25 00:00:00.000000000 Z
+date: 2020-05-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: digest-crc
@@ -489,9 +489,9 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: 2.1.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">="
+  - - ">"
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 1.3.1
 requirements: []
 rubyforge_project: 
 rubygems_version: 2.7.6