waterdrop 2.6.11 → 2.6.13

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3798d61ff3464dd6aaeb21f5457230a3b079748f260c59138142144efc62dc40
-  data.tar.gz: c9699bc202a67fbb8e456a4e0d316f56e8635e0986a09d0429a25999553c4a76
+  metadata.gz: '06984e8584d8b1ed3e6f854be6ca1cd1957acab63d6640c41087a42572da670a'
+  data.tar.gz: 9b0c94f047b5e21a554ecad5f32639a761aa32fbab9f5776a2470fdfe4adb675
 SHA512:
-  metadata.gz: c6a6d03d205314775b71ca0c4e922ef52159965e30b9280727712f25368f6af78c821d6d468b2362bd4094fbd2d4747195dc1e3a7cd76a0e4582f81a66829b43
-  data.tar.gz: 9541ae05afde65cf1124661fd57e23afb4265601748c6b4a60f6fc6f8cabe463cf59b6ac683e7d047e454cc671182808f800add0f68b780bbb1b1325abcab892
+  metadata.gz: 610ec69f42c6d209e3d024825d62ecd25e7d50ad212d14b4b2db7534463f63e5cee4662b0412db559d1537fa28ef76acd3b22418f686fcea6c6406a1cb6cec54
+  data.tar.gz: af35df96c8566d1fef3d16898553277c3d15a572ab8fc563858048beae70915a761109a32cd101b22123174f16dc601f7f28c2699fae20126d5dd9dc519fcd23

data/.github/workflows/ci.yml

@@ -1,6 +1,8 @@
 name: ci
 
-concurrency: ci-${{ github.ref }}
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
 
 on:
   pull_request:
@@ -16,13 +18,13 @@ jobs:
       fail-fast: false
       matrix:
         ruby:
-          - '3.3.0-preview2'
+          - '3.3'
           - '3.2'
           - '3.1'
           - '3.0'
           - '2.7'
         include:
-          - ruby: '3.2'
+          - ruby: '3.3'
             coverage: 'true'
     steps:
       - uses: actions/checkout@v4
@@ -35,6 +37,7 @@ jobs:
         with:
           ruby-version: ${{matrix.ruby}}
           bundler-cache: true
+          bundler: 'latest'
 
       - name: Run Kafka with docker-compose
         run: |
@@ -46,13 +49,25 @@ jobs:
 
       - name: Install latest bundler
         run: |
-          gem install bundler --no-document
+          if [[ "$(ruby -v | awk '{print $2}')" == 2.7.8* ]]; then
+            gem install bundler -v 2.4.22 --no-document
+            gem update --system 3.4.22 --no-document
+          else
+            gem install bundler --no-document
+            gem update --system --no-document
+          fi
+
           bundle config set without 'tools benchmarks docs'
 
       - name: Bundle install
         run: |
           bundle config set without development
-          bundle install --jobs 4 --retry 3
+
+          if [[ "$(ruby -v | awk '{print $2}')" == 2.7.8* ]]; then
+            BUNDLER_VERSION=2.4.22 bundle install --jobs 4 --retry 3
+          else
+            bundle install --jobs 4 --retry 3
+          fi
 
       - name: Run all tests
         env:
@@ -70,7 +85,7 @@ jobs:
       - name: Set up Ruby
         uses: ruby/setup-ruby@v1
         with:
-          ruby-version: 3.2
+          ruby-version: 3.3
       - name: Install latest bundler
         run: gem install bundler --no-document
       - name: Install Diffend plugin

data/.ruby-version

@@ -1 +1 @@
-3.2.2
+3.3.0

data/CHANGELOG.md

@@ -1,5 +1,14 @@
 # WaterDrop changelog
 
+## 2.6.13 (2024-01-29)
+- [Enhancement] Expose `#partition_count` for building custom partitioners that need to be aware of number of partitions on a given topic.
+
+## 2.6.12 (2024-01-03)
+- [Enhancement] Provide ability to label message dispatches for increased observability.
+- [Enhancement] Provide ability to commit offset during the transaction with a consumer provided.
+- [Change] Change transactional message purged error type from `message.error` to `librdkafka.dispatch_error` to align with the non-transactional error type.
+- [Change] Remove usage of concurrent ruby.
+
 ## 2.6.11 (2023-10-25)
 - [Enhancement] Return delivery handles and delivery report for both dummy and buffered clients with proper topics, partitions and offsets assign and auto-increment offsets per partition.
 - [Fix] Fix a case where buffered test client would not accumulate messages on failed transactions

data/Gemfile.lock

@@ -1,37 +1,48 @@
 PATH
   remote: .
   specs:
-    waterdrop (2.6.11)
+    waterdrop (2.6.13)
       karafka-core (>= 2.2.3, < 3.0.0)
       zeitwerk (~> 2.3)
 
 GEM
   remote: https://rubygems.org/
   specs:
-    activesupport (7.0.7.2)
+    activesupport (7.1.3)
+      base64
+      bigdecimal
       concurrent-ruby (~> 1.0, >= 1.0.2)
+      connection_pool (>= 2.2.5)
+      drb
       i18n (>= 1.6, < 2)
       minitest (>= 5.1)
+      mutex_m
       tzinfo (~> 2.0)
+    base64 (0.2.0)
+    bigdecimal (3.1.6)
     byebug (11.1.3)
-    concurrent-ruby (1.2.2)
+    concurrent-ruby (1.2.3)
+    connection_pool (2.4.1)
     diff-lcs (1.5.0)
     docile (1.4.0)
-    factory_bot (6.3.0)
+    drb (2.2.0)
+      ruby2_keywords
+    factory_bot (6.4.5)
       activesupport (>= 5.0.0)
-    ffi (1.15.5)
+    ffi (1.16.3)
     i18n (1.14.1)
       concurrent-ruby (~> 1.0)
-    karafka-core (2.2.3)
+    karafka-core (2.2.7)
       concurrent-ruby (>= 1.1)
-      karafka-rdkafka (>= 0.13.6, < 0.14.0)
-    karafka-rdkafka (0.13.6)
+      karafka-rdkafka (>= 0.13.9, < 0.15.0)
+    karafka-rdkafka (0.14.7)
       ffi (~> 1.15)
       mini_portile2 (~> 2.6)
       rake (> 12)
-    mini_portile2 (2.8.2)
-    minitest (5.19.0)
-    rake (13.0.6)
+    mini_portile2 (2.8.5)
+    minitest (5.21.2)
+    mutex_m (0.2.0)
+    rake (13.1.0)
     rspec (3.12.0)
       rspec-core (~> 3.12.0)
       rspec-expectations (~> 3.12.0)
@@ -41,10 +52,11 @@ GEM
     rspec-expectations (3.12.3)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.12.0)
-    rspec-mocks (3.12.5)
+    rspec-mocks (3.12.6)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.12.0)
-    rspec-support (3.12.0)
+    rspec-support (3.12.1)
+    ruby2_keywords (0.0.5)
     simplecov (0.22.0)
       docile (~> 1.1)
       simplecov-html (~> 0.11)
@@ -53,9 +65,10 @@ GEM
     simplecov_json_formatter (0.1.4)
     tzinfo (2.0.6)
       concurrent-ruby (~> 1.0)
-    zeitwerk (2.6.8)
+    zeitwerk (2.6.12)
 
 PLATFORMS
+  ruby
   x86_64-linux
 
 DEPENDENCIES
@@ -66,4 +79,4 @@ DEPENDENCIES
   waterdrop!
 
 BUNDLED WITH
-   2.4.12
+   2.5.4

data/config/locales/errors.yml

@@ -27,6 +27,12 @@ en:
       headers_invalid_key_type: all headers keys need to be of type String
       headers_invalid_value_type: all headers values need to be of type String
 
+    transactional_offset:
+      consumer_format: 'must respond to #consumer_group_metadata_pointer method'
+      message_format: 'must respond to #topic, #partition and #offset'
+      missing: must be present
+      offset_metadata_format: must be string or nil
+
     test:
       missing: must be present
       nested.id_format: 'is invalid'

data/docker-compose.yml

@@ -3,7 +3,7 @@ version: '2'
 services:
   kafka:
     container_name: kafka
-    image: confluentinc/cp-kafka:7.5.1
+    image: confluentinc/cp-kafka:7.5.3
 
     ports:
       - 9092:9092
@@ -23,3 +23,5 @@ services:
       KAFKA_AUTO_CREATE_TOPICS_ENABLE: 'true'
       KAFKA_TRANSACTION_STATE_LOG_REPLICATION_FACTOR: 1
       KAFKA_TRANSACTION_STATE_LOG_MIN_ISR: 1
+      KAFKA_ALLOW_EVERYONE_IF_NO_ACL_FOUND: "true"
+      KAFKA_AUTHORIZER_CLASS_NAME: org.apache.kafka.metadata.authorizer.StandardAuthorizer

data/lib/waterdrop/clients/buffered.rb

@@ -44,8 +44,6 @@ module WaterDrop
       def commit_transaction
         @transaction_level -= 1
 
-        return unless @transaction_level.zero?
-
         # Transfer transactional data on success
         @transaction_topics.each do |topic, messages|
           @topics[topic] += messages
@@ -58,12 +56,19 @@ module WaterDrop
         @transaction_active = false
       end
 
+      # Fakes storing the offset in a transactional fashion
+      #
+      # @param _consumer [#consumer_group_metadata_pointer] any consumer from which we can obtain
+      #   the librdkafka consumer group metadata pointer
+      # @param _tpl [Rdkafka::Consumer::TopicPartitionList] consumer tpl for offset storage
+      # @param _timeout [Integer] ms timeout
+      def send_offsets_to_transaction(_consumer, _tpl, _timeout)
+        nil
+      end
+
       # Aborts the transaction
       def abort_transaction
         @transaction_level -= 1
-
-        return unless @transaction_level.zero?
-
         @transaction_topics.clear
         @transaction_messages.clear
         @transaction_active = false

data/lib/waterdrop/contracts/transactional_offset.rb

@@ -0,0 +1,21 @@
+# frozen_string_literal: true
+
+module WaterDrop
+  module Contracts
+    # Contract to ensure that arguments provided to the transactional offset commit are valid
+    # and match our expectations
+    class TransactionalOffset < ::Karafka::Core::Contractable::Contract
+      configure do |config|
+        config.error_messages = YAML.safe_load(
+          File.read(
+            File.join(WaterDrop.gem_root, 'config', 'locales', 'errors.yml')
+          )
+        ).fetch('en').fetch('validations').fetch('transactional_offset')
+      end
+
+      required(:consumer) { |val| val.respond_to?(:consumer_group_metadata_pointer) }
+      required(:message) { |val| val.respond_to?(:topic) && val.respond_to?(:partition) }
+      required(:offset_metadata) { |val| val.is_a?(String) || val.nil? }
+    end
+  end
+end

data/lib/waterdrop/errors.rb

@@ -25,6 +25,9 @@ module WaterDrop
     # Raised when we want to send a message that is invalid (impossible topic, etc)
     MessageInvalidError = Class.new(BaseError)
 
+    # Raised when we want to commit transactional offset and the input is invalid
+    TransactionalOffsetInvalidError = Class.new(BaseError)
+
     # Raised when we've got an unexpected status. This should never happen. If it does, please
     # contact us as it is an error.
     StatusInvalidError = Class.new(BaseError)
@@ -32,7 +35,13 @@ module WaterDrop
     # Raised when there is an inline error during single message produce operations
     ProduceError = Class.new(BaseError)
 
+    # Raised when we attempt to perform operation that is only allowed inside of a transaction and
+    # there is no transaction around us
+    TransactionRequiredError = Class.new(BaseError)
+
     # Raise it within a transaction to abort it
+    # It does not have an `Error` postfix because technically it is not an error as it is used for
+    # graceful transaction aborting
     AbortTransaction = Class.new(BaseError)
 
     # Raised when during messages producing something bad happened inline

data/lib/waterdrop/helpers/counter.rb

@@ -0,0 +1,27 @@
+# frozen_string_literal: true
+
+module WaterDrop
+  # Extra internal helper objects
+  module Helpers
+    # Atomic counter that we can safely increment and decrement without race conditions
+    class Counter
+      # @return [Integer] current value
+      attr_reader :value
+
+      def initialize
+        @value = 0
+        @mutex = Mutex.new
+      end
+
+      # Increments the value by 1
+      def increment
+        @mutex.synchronize { @value += 1 }
+      end
+
+      # Decrements the value by 1
+      def decrement
+        @mutex.synchronize { @value -= 1 }
+      end
+    end
+  end
+end

data/lib/waterdrop/instrumentation/callbacks/delivery.rb

@@ -57,7 +57,8 @@ module WaterDrop
             offset: delivery_report.offset,
             partition: delivery_report.partition,
             topic: delivery_report.topic_name,
-            delivery_report: delivery_report
+            delivery_report: delivery_report,
+            label: delivery_report.label
           )
         end
 
@@ -71,7 +72,9 @@ module WaterDrop
             offset: delivery_report.offset,
             partition: delivery_report.partition,
             topic: delivery_report.topic_name,
-            delivery_report: delivery_report
+            delivery_report: delivery_report,
+            label: delivery_report.label,
+            type: 'librdkafka.dispatch_error'
           )
         end
 
@@ -86,6 +89,7 @@ module WaterDrop
             partition: delivery_report.partition,
             topic: delivery_report.topic_name,
             delivery_report: delivery_report,
+            label: delivery_report.label,
             type: 'librdkafka.dispatch_error'
           )
         end

data/lib/waterdrop/instrumentation/logger_listener.rb

@@ -145,6 +145,20 @@ module WaterDrop
         info(event, 'Committing transaction')
       end
 
+      # @param event [Dry::Events::Event] event that happened with the details
+      def on_transaction_marked_as_consumed(event)
+        message = event[:message]
+        topic = message.topic
+        partition = message.partition
+        offset = message.offset
+        loc = "#{topic}/#{partition}"
+
+        info(
+          event,
+          "Marking message with offset #{offset} for topic #{loc} as consumed in a transaction"
+        )
+      end
+
       # @param event [Dry::Events::Event] event that happened with the details
       def on_transaction_finished(event)
         info(event, 'Processing transaction')

data/lib/waterdrop/instrumentation/notifications.rb

@@ -22,6 +22,7 @@ module WaterDrop
         transaction.started
         transaction.committed
         transaction.aborted
+        transaction.marked_as_consumed
         transaction.finished
 
         buffer.flushed_async

data/lib/waterdrop/producer/buffer.rb

@@ -82,7 +82,7 @@ module WaterDrop
 
         @buffer_mutex.synchronize do
           data_for_dispatch = @messages
-          @messages = Concurrent::Array.new
+          @messages = []
         end
 
         # Do nothing if nothing to flush

data/lib/waterdrop/producer/transactions.rb

@@ -4,6 +4,11 @@ module WaterDrop
   class Producer
     # Transactions related producer functionalities
     module Transactions
+      # Contract to validate that input for transactional offset storage is correct
+      CONTRACT = Contracts::TransactionalOffset.new
+
+      private_constant :CONTRACT
+
       # Creates a transaction.
       #
       # Karafka transactions work in a similar manner to SQL db transactions though there are some
@@ -91,6 +96,49 @@ module WaterDrop
         @transactional = config.kafka.to_h.key?(:'transactional.id')
       end
 
+      # Marks given message as consumed inside of a transaction.
+      #
+      # @param consumer [#consumer_group_metadata_pointer] any consumer from which we can obtain
+      #   the librdkafka consumer group metadata pointer
+      # @param message [Karafka::Messages::Message] karafka message
+      # @param offset_metadata [String] offset metadata or nil if none
+      def transaction_mark_as_consumed(consumer, message, offset_metadata = nil)
+        raise Errors::TransactionRequiredError unless @transaction_mutex.owned?
+
+        CONTRACT.validate!(
+          {
+            consumer: consumer,
+            message: message,
+            offset_metadata: offset_metadata
+          },
+          Errors::TransactionalOffsetInvalidError
+        )
+
+        details = { message: message, offset_metadata: offset_metadata }
+
+        transactional_instrument(:marked_as_consumed, details) do
+          tpl = Rdkafka::Consumer::TopicPartitionList.new
+          partition = Rdkafka::Consumer::Partition.new(
+            message.partition,
+            # +1 because this is next offset from which we will start processing from
+            message.offset + 1,
+            0,
+            offset_metadata
+          )
+
+          tpl.add_topic_and_partitions_with_offsets(message.topic, [partition])
+
+          with_transactional_error_handling(:store_offset) do
+            client.send_offsets_to_transaction(
+              consumer,
+              tpl,
+              # This setting is at the moment in seconds and we require ms
+              @config.max_wait_timeout * 1_000
+            )
+          end
+        end
+      end
+
       private
 
       # Runs provided code with a transaction wrapper if transactions are enabled.
@@ -105,9 +153,10 @@ module WaterDrop
       # Instruments the transactional operation with producer id
       #
       # @param key [Symbol] transaction operation key
+      # @param details [Hash] additional instrumentation details
       # @param block [Proc] block to run inside the instrumentation or nothing if not given
-      def transactional_instrument(key, &block)
-        @monitor.instrument("transaction.#{key}", producer_id: id, &block)
+      def transactional_instrument(key, details = EMPTY_HASH, &block)
+        @monitor.instrument("transaction.#{key}", details.merge(producer_id: id), &block)
       end
 
       # Error handling for transactional operations is a bit special. There are three types of
@@ -157,7 +206,7 @@ module WaterDrop
           # Always attempt to abort but if aborting fails with an abortable error, do not attempt
           # to abort from abort as this could create an infinite loop
           with_transactional_error_handling(:abort, allow_abortable: false) do
-            transactional_instrument(:aborted) { @client.abort_transaction }
+            transactional_instrument(:aborted) { client.abort_transaction }
           end
 
           raise

data/lib/waterdrop/producer.rb

@@ -16,7 +16,10 @@ module WaterDrop
       Rdkafka::Producer::DeliveryHandle::WaitTimeoutError
     ].freeze
 
-    private_constant :SUPPORTED_FLOW_ERRORS
+    # Empty has to save on memory allocations
+    EMPTY_HASH = {}.freeze
+
+    private_constant :SUPPORTED_FLOW_ERRORS, :EMPTY_HASH
 
     def_delegators :config, :middleware
 
@@ -24,7 +27,7 @@ module WaterDrop
     attr_reader :id
     # @return [Status] producer status object
     attr_reader :status
-    # @return [Concurrent::Array] internal messages buffer
+    # @return [Array] internal messages buffer
     attr_reader :messages
     # @return [Object] monitor we want to use
     attr_reader :monitor
@@ -35,14 +38,14 @@ module WaterDrop
     # @param block [Proc] configuration block
     # @return [Producer] producer instance
     def initialize(&block)
-      @operations_in_progress = Concurrent::AtomicFixnum.new(0)
+      @operations_in_progress = Helpers::Counter.new
       @buffer_mutex = Mutex.new
       @connecting_mutex = Mutex.new
       @operating_mutex = Mutex.new
       @transaction_mutex = Mutex.new
 
       @status = Status.new
-      @messages = Concurrent::Array.new
+      @messages = []
 
       return unless block
 
@@ -119,6 +122,16 @@ module WaterDrop
       @client
     end
 
+    # Fetches and caches the partition count of a topic
+    #
+    # @param topic [String] topic for which we want to get the number of partitions
+    # @return [Integer] number of partitions of the requested topic
+    #
+    # @note It uses the underlying `rdkafka-ruby` partition count cache.
+    def partition_count(topic)
+      client.partition_count(topic.to_s)
+    end
+
     # Purges data from both the buffer queue as well as the librdkafka queue.
     #
     # @note This is an operation that can cause data loss. Keep that in mind. It will not only
@@ -127,7 +140,7 @@ module WaterDrop
     def purge
       @monitor.instrument('buffer.purged', producer_id: id) do
         @buffer_mutex.synchronize do
-          @messages = Concurrent::Array.new
+          @messages = []
         end
 
         @client.purge

data/lib/waterdrop/version.rb

@@ -3,5 +3,5 @@
 # WaterDrop library
 module WaterDrop
   # Current WaterDrop version
-  VERSION = '2.6.11'
+  VERSION = '2.6.13'
 end

data/lib/waterdrop.rb

@@ -9,7 +9,6 @@
   securerandom
   karafka-core
   pathname
-  concurrent/atomic/atomic_fixnum
 ].each { |lib| require lib }
 
 # WaterDrop library

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: waterdrop
 version: !ruby/object:Gem::Version
-  version: 2.6.11
+  version: 2.6.13
 platform: ruby
 authors:
 - Maciej Mensfeld
@@ -35,7 +35,7 @@ cert_chain:
   AnG1dJU+yL2BK7vaVytLTstJME5mepSZ46qqIJXMuWob/YPDmVaBF39TDSG9e34s
   msG3BiCqgOgHAnL23+CN3Rt8MsuRfEtoTKpJVcCfoEoNHOkc
   -----END CERTIFICATE-----
-date: 2023-10-25 00:00:00.000000000 Z
+date: 2024-01-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: karafka-core
@@ -102,7 +102,9 @@ files:
 - lib/waterdrop/contracts.rb
 - lib/waterdrop/contracts/config.rb
 - lib/waterdrop/contracts/message.rb
+- lib/waterdrop/contracts/transactional_offset.rb
 - lib/waterdrop/errors.rb
+- lib/waterdrop/helpers/counter.rb
 - lib/waterdrop/instrumentation/callbacks/delivery.rb
 - lib/waterdrop/instrumentation/callbacks/error.rb
 - lib/waterdrop/instrumentation/callbacks/statistics.rb
@@ -149,7 +151,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.4.19
+rubygems_version: 3.5.3
 signing_key: 
 specification_version: 4
 summary: Kafka messaging made easy!