fluent-plugin-kafka 0.17.5 → 0.18.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: eb7f693666ff7fd4bdb43b249d3e60e882ebf99e80402de2a5310b61668ef9e7
-  data.tar.gz: d4b141409cf83402ad1e28b9579317c672b20c9760b8f58b8d54800e8c46313b
+  metadata.gz: 0a8590f6d34bbdb2faa991bba6e32a1424623c7ddd9609dc1d2ffcbeabaa20e1
+  data.tar.gz: a653d9dba00fa82f18071304be9093cd6d620bf4bda29d5e00a75113d05e8e3e
 SHA512:
-  metadata.gz: 8f86c7c82fbf5db63ef0c92737ba4935a3a60534d434dfb9552c2ad624ec1e28fea9876c689a29ca6b16c16abc3818fdfb24bd7540495f980daa8a5509b29ca0
-  data.tar.gz: 8b1dfba11d40e12f9b0b1fbf4ff947678fb5ef2ab3fbd3969e363935c8e895c29f497273ddb86c80c4e1ebc9d254045275dc7e08b07788d4573799c97ce7a393
+  metadata.gz: 69a784e7d69e2229a036110997b2691143008d4ca786e18415e9a806a6d577b38465a4a3931166418a5b7e022dddbc62d014c0c784277ccaaecc9b6570822564
+  data.tar.gz: 9f14d8d10a45bf390c801c93d79b893f4248ded7e37dc87f8c4453b04158754e1006aa7603874986e8c93812fe899197219abccb3bdb0ae4daacef6d516a0af3

data/ChangeLog

@@ -1,3 +1,8 @@
+Release 0.18.0 - 2022/07/21
+	* out_kafka2: Keep alive Kafka connections between flushes
+	* out_rdkafka2: Enable to set SASL credentials via `username` and `password` parameters
+	* out_kafka2/out_rdkafka2: Add `record_key` parameter
+
 Release 0.17.5 - 2022/03/18
 	* out_kafka2: Add `resolve_seed_brokers` parameter
 

data/README.md

@@ -193,6 +193,7 @@ If `ruby-kafka` doesn't fit your kafka environment, check `rdkafka2` plugin inst
       message_key_key       (string) :default => 'message_key'
       default_topic         (string) :default => nil
       default_partition_key (string) :default => nil
+      record_key            (string) :default => nil
       default_message_key   (string) :default => nil
       exclude_topic_key     (bool)   :default => false
       exclude_partition_key (bool)   :default => false
@@ -205,6 +206,7 @@ If `ruby-kafka` doesn't fit your kafka environment, check `rdkafka2` plugin inst
       use_default_for_unknown_topic (bool) :default => false
       discard_kafka_delivery_failed (bool) :default => false (No discard)
       partitioner_hash_function (enum) (crc32|murmur2) :default => 'crc32'
+      share_producer        (bool)   :default => false
 
       <format>
         @type (json|ltsv|msgpack|attr:<record name>|<formatter name>) :default => json
@@ -335,6 +337,40 @@ For example, `$.source.ip` can be extracted with config `headers_from_record` an
 
 > Using this config to remove unused fields is discouraged. A [filter plugin](https://docs.fluentd.org/v/0.12/filter) can be used for this purpose.
 
+#### Send only a sub field as a message payload
+
+If `record_key` is provided, the plugin sends only a sub field given by that key.
+The configuration format is jsonpath.
+
+e.g. When the following configuration and the incoming record are given:
+
+configuration:
+
+    <match **>
+      @type kafka2
+      [...]
+      record_key '$.data'
+    </match>
+
+record:
+
+    {
+        "specversion" : "1.0",
+        "type" : "com.example.someevent",
+        "id" : "C234-1234-1234",
+        "time" : "2018-04-05T17:31:00Z",
+        "datacontenttype" : "application/json",
+        "data" : {
+            "appinfoA" : "abc",
+            "appinfoB" : 123,
+            "appinfoC" : true
+        },
+        ...
+    }
+
+only the `data` field will be serialized by the formatter and sent to Kafka.
+The toplevel `data` key will be removed.
+
 ### Buffered output plugin
 
 This plugin uses ruby-kafka producer for writing data. This plugin is for v0.12. If you use v1, see `kafka2`.
@@ -460,6 +496,7 @@ You need to install rdkafka gem.
       # same with kafka2
       headers               (hash) :default => {}
       headers_from_record   (hash) :default => {}
+      record_key            (string) :default => nil
 
       <format>
         @type (json|ltsv|msgpack|attr:<record name>|<formatter name>) :default => json

data/fluent-plugin-kafka.gemspec

@@ -13,7 +13,7 @@ Gem::Specification.new do |gem|
   gem.test_files    = gem.files.grep(%r{^(test|spec|features)/})
   gem.name          = "fluent-plugin-kafka"
   gem.require_paths = ["lib"]
-  gem.version       = '0.17.5'
+  gem.version       = '0.18.0'
   gem.required_ruby_version = ">= 2.1.0"
 
   gem.add_dependency "fluentd", [">= 0.10.58", "< 2"]

data/lib/fluent/plugin/out_kafka2.rb

@@ -27,6 +27,11 @@ DESC
     config_param :partitioner_hash_function, :enum, list: [:crc32, :murmur2], :default => :crc32,
                  :desc => "Specify kafka patrtitioner hash algorithm"
     config_param :default_partition, :integer, :default => nil
+    config_param :record_key, :string, :default => nil,
+                 :desc => <<-DESC
+A jsonpath to a record value pointing to the field which will be passed to the formatter and sent as the Kafka message payload.
+If defined, only this field in the record will be sent to Kafka as the message payload.
+DESC
     config_param :use_default_for_unknown_topic, :bool, :default => false, :desc => "If true, default_topic is used when topic not found"
     config_param :client_id, :string, :default => 'fluentd'
     config_param :idempotent, :bool, :default => false, :desc => 'Enable idempotent producer'
@@ -81,6 +86,7 @@ DESC
 Add a regular expression to capture ActiveSupport notifications from the Kafka client
 requires activesupport gem - records will be generated under fluent_kafka_stats.**
 DESC
+    config_param :share_producer, :bool, :default => false, :desc => 'share kafka producer between flush threads'
 
     config_section :buffer do
       config_set_default :chunk_keys, ["topic"]
@@ -96,6 +102,12 @@ DESC
       super
 
       @kafka = nil
+      @producers = nil
+      @producers_mutex = nil
+      @shared_producer = nil
+
+      @writing_threads_mutex = Mutex.new
+      @writing_threads = Set.new
     end
 
     def refresh_client(raise_error = true)
@@ -185,15 +197,29 @@ DESC
       @exclude_field_accessors = @exclude_fields.map do |field|
         record_accessor_create(field)
       end
+
+      @record_field_accessor = nil
+      @record_field_accessor = record_accessor_create(@record_key) unless @record_key.nil?
     end
 
     def multi_workers_ready?
       true
     end
 
+    def create_producer
+      @kafka.producer(**@producer_opts)
+    end
+
     def start
       super
       refresh_client
+
+      if @share_producer
+        @shared_producer = create_producer
+      else
+        @producers = {}
+        @producers_mutex = Mutex.new
+      end
     end
 
     def close
@@ -206,6 +232,56 @@ DESC
       @kafka = nil
     end
 
+    def wait_writing_threads
+      done = false
+      until done do
+        @writing_threads_mutex.synchronize do
+          done = true if @writing_threads.empty?
+        end
+        sleep(1) unless done
+      end
+    end
+
+    def shutdown
+      super
+      wait_writing_threads
+      shutdown_producers
+    end
+
+    def shutdown_producers
+      if @share_producer
+        @shared_producer.shutdown
+        @shared_producer = nil
+      else
+        @producers_mutex.synchronize {
+          shutdown_threads = @producers.map { |key, producer|
+            th = Thread.new {
+              producer.shutdown
+            }
+            th.abort_on_exception = true
+            th
+          }
+          shutdown_threads.each { |th| th.join }
+          @producers = {}
+        }
+      end
+    end
+
+    def get_producer
+      if @share_producer
+        @shared_producer
+      else
+        @producers_mutex.synchronize {
+          producer = @producers[Thread.current.object_id]
+          unless producer
+            producer = create_producer
+            @producers[Thread.current.object_id] = producer
+          end
+          producer
+        }
+      end
+    end
+
     def setup_formatter(conf)
       type = conf['@type']
       case type
@@ -229,6 +305,8 @@ DESC
 
     # TODO: optimize write performance
     def write(chunk)
+      @writing_threads_mutex.synchronize { @writing_threads.add(Thread.current) }
+
       tag = chunk.metadata.tag
       topic = if @topic
                 extract_placeholders(@topic, chunk)
@@ -237,13 +315,12 @@ DESC
               end
 
       messages = 0
-      record_buf = nil
 
       base_headers = @headers
       mutate_headers = !@headers_from_record_accessors.empty?
 
       begin
-        producer = @kafka.topic_producer(topic, **@producer_opts)
+        producer = get_producer
         chunk.msgpack_each { |time, record|
           begin
             record = inject_values_to_record(tag, time, record)
@@ -267,6 +344,7 @@ DESC
               end
             end
 
+            record = @record_field_accessor.call(record) unless @record_field_accessor.nil?
             record_buf = @formatter_proc.call(tag, time, record)
             record_buf_bytes = record_buf.bytesize
             if @max_send_limit_bytes && record_buf_bytes > @max_send_limit_bytes
@@ -283,7 +361,7 @@ DESC
           messages += 1
 
           producer.produce(record_buf, key: message_key, partition_key: partition_key, partition: partition, headers: headers,
-                           create_time: @use_event_time ? Time.at(time) : Time.now)
+                           create_time: @use_event_time ? Time.at(time) : Time.now, topic: topic)
         }
 
         if messages > 0
@@ -301,7 +379,6 @@ DESC
         end
       rescue Kafka::UnknownTopicOrPartition
         if @use_default_for_unknown_topic && topic != @default_topic
-          producer.shutdown if producer
           log.warn "'#{topic}' topic not found. Retry with '#{default_topic}' topic"
           topic = @default_topic
           retry
@@ -321,7 +398,7 @@ DESC
       # Raise exception to retry sendind messages
       raise e unless ignore
     ensure
-      producer.shutdown if producer
+      @writing_threads_mutex.synchronize { @writing_threads.delete(Thread.current) }
     end
   end
 end

data/lib/fluent/plugin/out_rdkafka2.rb

@@ -73,6 +73,11 @@ DESC
                  :desc => <<-DESC
 The codec the producer uses to compress messages. Used for compression.codec
 Supported codecs: (gzip|snappy)
+DESC
+    config_param :record_key, :string, :default => nil,
+                 :desc => <<-DESC
+A jsonpath to a record value pointing to the field which will be passed to the formatter and sent as the Kafka message payload.
+If defined, only this field in the record will be sent to Kafka as the message payload.
 DESC
     config_param :use_event_time, :bool, :default => false, :desc => 'Use fluentd event time for rdkafka timestamp'
     config_param :max_send_limit_bytes, :size, :default => nil
@@ -230,6 +235,9 @@ DESC
       end
 
       @enqueue_rate = EnqueueRate.new(@max_enqueue_bytes_per_second) unless @max_enqueue_bytes_per_second.nil?
+
+      @record_field_accessor = nil
+      @record_field_accessor = record_accessor_create(@record_key) unless @record_key.nil?
     end
 
     def build_config
@@ -270,6 +278,8 @@ DESC
       config[:"queue.buffering.max.messages"] = @rdkafka_buffering_max_messages if @rdkafka_buffering_max_messages
       config[:"message.max.bytes"] = @rdkafka_message_max_bytes if @rdkafka_message_max_bytes
       config[:"batch.num.messages"] = @rdkafka_message_max_num if @rdkafka_message_max_num
+      config[:"sasl.username"] = @username if @username
+      config[:"sasl.password"] = @password if @password
 
       @rdkafka_options.each { |k, v|
         config[k.to_sym] = v
@@ -371,8 +381,6 @@ DESC
               end
 
       handlers = []
-      record_buf = nil
-      record_buf_bytes = nil
 
       headers = @headers.clone
 
@@ -395,6 +403,7 @@ DESC
               end
             end
 
+            record = @record_field_accessor.call(record) unless @record_field_accessor.nil?
             record_buf = @formatter_proc.call(tag, time, record)
             record_buf_bytes = record_buf.bytesize
             if @max_send_limit_bytes && record_buf_bytes > @max_send_limit_bytes

data/test/plugin/test_out_kafka2.rb

@@ -105,6 +105,21 @@ class Kafka2OutputTest < Test::Unit::TestCase
       assert_equal([expected_message], actual_messages)
     end
 
+    def test_record_key
+      conf = config(default_topic: TOPIC_NAME) +
+        config_element('ROOT', '', {"record_key" => "$.data"}, [])
+      target_driver = create_target_driver
+      target_driver.run(expect_records: 1, timeout: 5) do
+        sleep 2
+        d = create_driver(conf)
+        d.run do
+          d.feed('test', event_time, {'data' => {'a' => 'b', 'foo' => 'bar', 'message' => 'test'}, 'message_key' => '123456'})
+        end
+      end
+      actual_messages = target_driver.events.collect { |event| event[2] }
+      assert_equal([{'a' => 'b', 'foo' => 'bar', 'message' => 'test'}], actual_messages)
+    end
+
     def test_exclude_fields
       conf = config(default_topic: TOPIC_NAME) +
              config_element('ROOT', '', {"exclude_fields" => "$.foo"}, [])

data/test/plugin/test_out_rdkafka2.rb

@@ -163,5 +163,20 @@ class Rdkafka2OutputTest < Test::Unit::TestCase
       actual_messages = target_driver.events.collect { |event| event[2] }
       assert_equal(expected_messages, actual_messages)
     end
+
+    def test_record_key
+      conf = config(default_topic: TOPIC_NAME) +
+             config_element('ROOT', '', {"record_key" => "$.data"}, [])
+      target_driver = create_target_driver
+      target_driver.run(expect_records: 1, timeout: 5) do
+        sleep 2
+        d = create_driver(conf)
+        d.run do
+          d.feed('test', event_time, {'data' => {'a' => 'b', 'foo' => 'bar', 'message' => 'test'}, 'message_key' => '123456'})
+        end
+      end
+      actual_messages = target_driver.events.collect { |event| event[2] }
+      assert_equal([{'a' => 'b', 'foo' => 'bar', 'message' => 'test'}], actual_messages)
+    end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: fluent-plugin-kafka
 version: !ruby/object:Gem::Version
-  version: 0.17.5
+  version: 0.18.0
 platform: ruby
 authors:
 - Hidemasa Togashi
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2022-03-18 00:00:00.000000000 Z
+date: 2022-07-21 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: fluentd
@@ -193,7 +193,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.2.5
+rubygems_version: 3.3.5
 signing_key: 
 specification_version: 4
 summary: Fluentd plugin for Apache Kafka > 0.8