fluent-plugin-webhdfs 1.2.4 → 1.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: fcefccfd606717624933c095b4450489b3a24bd40cebf8af180e824e77ece775
-  data.tar.gz: 2d307f89de1c026443bcef00b7e821d884b8df2d64a79c37561617cc74ca6a5e
+  metadata.gz: 2d96f9304470f4d3409a1209b96c56722a62acfefc55633d81525e21103ec7e9
+  data.tar.gz: 8c6982670e30e112815a3abec28a2865609e4ca2cccbe097a42b6abb9080af21
 SHA512:
-  metadata.gz: f318228d5b577be1151fa78951e317c68c862ef8981425d148d55d2e4df648004638fe8373ce6e7028553b6f06143e4dda8cc2318b27223090be9e454c88c5ac
-  data.tar.gz: 7ff49536fe66c00f4888bb2266e78c9f58534e3789891c41db1c9a33300fcd3218966cb3d8e88398dcdca4a2e41136ced6f1452a7e9d095d12297a5ddd07a842
+  metadata.gz: a56a3b8ac2e7bf279ddb23d5a4fafb187289883442b01355f86cf3d626332aff35a78f4fee38c32767a7162741b550143fbfc50f9d953db0ce8a8220d022d35f
+  data.tar.gz: 7110d25391fc90d0e0aa8042014596994b3fe700737a08febd1f1d69485fd9eef212f0882e51832e2a3f95d64a7e21e0091f8f8e37444151404104230645029d

data/.gitignore

@@ -15,3 +15,4 @@ spec/reports
 test/tmp
 test/version_tmp
 tmp
+vendor/

data/.travis.yml

@@ -2,10 +2,11 @@ sudo: false
 language: ruby
 
 rvm:
-  - 2.1
-  - 2.2
-  - 2.3.3
-  - 2.4.0
+  - 2.4
+  - 2.5
+  - 2.6
+  - 2.7
+  - ruby-head
 
 branches:
   only:
@@ -23,3 +24,7 @@ script: bundle exec rake test
 
 gemfile:
   - Gemfile
+
+matrix:
+  allow_failures:
+    - rvm: ruby-head

data/README.md

@@ -157,14 +157,43 @@ If you want to compress data before storing it:
       host namenode.your.cluster.local
       port 50070
       path /path/on/hdfs/access.log.%Y%m%d_%H
-      compress gzip  # or 'bzip2', 'snappy', 'lzo_command'
+      compress gzip  # or 'bzip2', 'snappy', 'hadoop_snappy', 'lzo_command', 'zstd'
     </match>
 
-Note that if you set `compress gzip`, then the suffix `.gz` will be added to path (or `.bz2`, `sz`, `.lzo`).
+Note that if you set `compress gzip`, then the suffix `.gz` will be added to path (or `.bz2`, `.sz`, `.snappy`, `.lzo`, `.zst`).
 Note that you have to install additional gem for several compress algorithms:
 
 - snappy: install snappy gem
+- hadoop_snappy: install snappy gem
 - bzip2: install bzip2-ffi gem
+- zstd: install zstandard gem
+
+Note that zstd will require installation of the libzstd native library. See the [zstandard-ruby](https://github.com/msievers/zstandard-ruby#examples-for-installing-libzstd) repo for infomration on the required packages for your operating system.
+
+You can also specify compression block size (currently supported only for Snappy codecs):
+
+    <match access.**>
+      @type webhdfs
+      host namenode.your.cluster.local
+      port 50070
+      path /path/on/hdfs/access.log.%Y%m%d_%H
+      compress hadoop_snappy
+      block_size 32768
+    </match>
+
+If you want to explicitly specify file extensions in HDFS (override default compressor extensions):
+
+    <match access.**>
+      @type webhdfs
+      host namenode.your.cluster.local
+      port 50070
+      path /path/on/hdfs/access.log.%Y%m%d_%H
+      compress snappy
+      extension ".snappy"
+    </match>
+
+With this configuration paths in HDFS will be like `/path/on/hdfs/access.log.20201003_12.snappy`.
+This one may be useful when (for example) you need to use snappy codec but `.sz` files are not recognized as snappy files in HDFS.
 
 ### Namenode HA / Auto retry for WebHDFS known errors
 

data/fluent-plugin-webhdfs.gemspec

@@ -2,7 +2,7 @@
 
 Gem::Specification.new do |gem|
   gem.name          = "fluent-plugin-webhdfs"
-  gem.version       = "1.2.4"
+  gem.version       = "1.4.0"
   gem.authors       = ["TAGOMORI Satoshi"]
   gem.email         = ["tagomoris@gmail.com"]
   gem.summary       = %q{Fluentd plugin to write data on HDFS over WebHDFS, with flexible formatting}
@@ -21,6 +21,7 @@ Gem::Specification.new do |gem|
   gem.add_development_dependency "appraisal"
   gem.add_development_dependency "snappy", '>= 0.0.13'
   gem.add_development_dependency "bzip2-ffi"
+  gem.add_development_dependency "zstandard"
   gem.add_runtime_dependency "fluentd", '>= 0.14.22'
   gem.add_runtime_dependency "webhdfs", '>= 0.6.0'
 end

data/lib/fluent/plugin/out_webhdfs.rb

@@ -67,10 +67,13 @@ class Fluent::Plugin::WebHDFSOutput < Fluent::Plugin::Output
   desc 'kerberos keytab file'
   config_param :kerberos_keytab, :string, default: nil
 
-  SUPPORTED_COMPRESS = [:gzip, :bzip2, :snappy, :lzo_command, :text]
-  desc "Compress method (#{SUPPORTED_COMPRESS.join(',')})"
+  SUPPORTED_COMPRESS = [:gzip, :bzip2, :snappy, :hadoop_snappy, :lzo_command, :zstd, :text]
+  desc "Compression method (#{SUPPORTED_COMPRESS.join(',')})"
   config_param :compress, :enum, list: SUPPORTED_COMPRESS, default: :text
 
+  desc 'HDFS file extensions (overrides default compressor extensions)'
+  config_param :extension, :string, default: nil
+
   config_param :remove_prefix, :string, default: nil, deprecated: "use @label for routing"
   config_param :default_tag, :string, default: nil, deprecated: "use @label for routing"
   config_param :null_value, :string, default: nil, deprecated: "use filter plugins to convert null values into any specified string"
@@ -153,6 +156,7 @@ class Fluent::Plugin::WebHDFSOutput < Fluent::Plugin::Output
     end
 
     @compressor = COMPRESSOR_REGISTRY.lookup(@compress.to_s).new
+    @compressor.configure(conf)
 
     if @host
       @namenode_host = @host
@@ -267,17 +271,22 @@ class Fluent::Plugin::WebHDFSOutput < Fluent::Plugin::Output
   end
 
   def send_data(path, data)
-    if @append
-      begin
-        @client.append(path, data)
-      rescue WebHDFS::FileNotFoundError
-        @client.create(path, data)
-      end
+    return @client.create(path, data, {'overwrite' => 'true'}) unless @append
+
+    if path_exists?(path)
+      @client.append(path, data)
     else
-      @client.create(path, data, {'overwrite' => 'true'})
+      @client.create(path, data)
     end
   end
 
+  def path_exists?(path)
+    @client.stat(path)
+    true
+  rescue WebHDFS::FileNotFoundError
+    false
+  end
+
   HOSTNAME_PLACEHOLDERS_DEPRECATED = ['${hostname}', '%{hostname}', '__HOSTNAME__']
   UUID_RANDOM_PLACEHOLDERS_DEPRECATED = ['${uuid}', '${uuid:random}', '__UUID__', '__UUID_RANDOM__']
   UUID_OTHER_PLACEHOLDERS_OBSOLETED = ['${uuid:hostname}', '%{uuid:hostname}', '__UUID_HOSTNAME__', '${uuid:timestamp}', '%{uuid:timestamp}', '__UUID_TIMESTAMP__']
@@ -319,7 +328,8 @@ class Fluent::Plugin::WebHDFSOutput < Fluent::Plugin::Output
                 else
                   extract_placeholders(@path.gsub(CHUNK_ID_PLACE_HOLDER, dump_unique_id_hex(chunk.unique_id)), chunk)
                 end
-    hdfs_path = "#{hdfs_path}#{@compressor.ext}"
+    hdfs_ext = @extension || @compressor.ext
+    hdfs_path = "#{hdfs_path}#{hdfs_ext}"
     if @replace_random_uuid
       uuid_random = SecureRandom.uuid
       hdfs_path = hdfs_path.gsub('%{uuid}', uuid_random).gsub('%{uuid_flush}', uuid_random)
@@ -502,7 +512,7 @@ class Fluent::Plugin::WebHDFSOutput < Fluent::Plugin::Output
       begin
         Open3.capture3("#{command} -V")
       rescue Errno::ENOENT
-        raise ConfigError, "'#{command}' utility must be in PATH for #{algo} compression"
+        raise Fluent::ConfigError, "'#{command}' utility must be in PATH for #{algo} compression"
       end
     end
   end
@@ -518,4 +528,6 @@ require 'fluent/plugin/webhdfs_compressor_text'
 require 'fluent/plugin/webhdfs_compressor_gzip'
 require 'fluent/plugin/webhdfs_compressor_bzip2'
 require 'fluent/plugin/webhdfs_compressor_snappy'
+require 'fluent/plugin/webhdfs_compressor_hadoop_snappy'
 require 'fluent/plugin/webhdfs_compressor_lzo_command'
+require 'fluent/plugin/webhdfs_compressor_zstd'

data/lib/fluent/plugin/webhdfs_compressor_hadoop_snappy.rb

@@ -0,0 +1,32 @@
+module Fluent::Plugin
+  class WebHDFSOutput < Output
+    class HadoopSnappyCompressor < Compressor
+      WebHDFSOutput.register_compressor('hadoop_snappy', self)
+
+      DEFAULT_BLOCK_SIZE = 256 * 1024
+
+      desc 'Block size for compression algorithm'
+      config_param :block_size, :integer, default: DEFAULT_BLOCK_SIZE
+
+      def initialize(options = {})
+        super()
+        begin
+          require "snappy"
+        rescue LoadError
+          raise Fluent::ConfigError, "Install snappy before using snappy compressor"
+        end
+      end
+
+      def ext
+        ".snappy"
+      end
+
+      def compress(chunk, tmp)
+        Snappy::Hadoop::Writer.new(tmp, @block_size) do |w|
+          w << chunk.read
+          w.flush
+        end
+      end
+    end
+  end
+end

data/lib/fluent/plugin/webhdfs_compressor_lzo_command.rb

@@ -5,8 +5,8 @@ module Fluent::Plugin
 
       config_param :command_parameter, :string, default: '-qf1'
 
-      def configure(conf)
-        super
+      def initialize(options = {})
+        super()
         check_command('lzop', 'LZO')
       end
 

data/lib/fluent/plugin/webhdfs_compressor_snappy.rb

@@ -3,11 +3,17 @@ module Fluent::Plugin
     class SnappyCompressor < Compressor
       WebHDFSOutput.register_compressor('snappy', self)
 
+      DEFAULT_BLOCK_SIZE = 32 * 1024
+
+      desc 'Block size for compression algorithm'
+      config_param :block_size, :integer, default: DEFAULT_BLOCK_SIZE
+
       def initialize(options = {})
+        super()
         begin
           require "snappy"
         rescue LoadError
-          raise Fluent::ConfigError, "Install snappy before use snappy compressor"
+          raise Fluent::ConfigError, "Install snappy before using snappy compressor"
         end
       end
 
@@ -16,9 +22,10 @@ module Fluent::Plugin
       end
 
       def compress(chunk, tmp)
-        w = Snappy::Writer.new(tmp)
-        chunk.write_to(w)
-        w.close
+        Snappy::Writer.new(tmp, @block_size) do |w|
+          w << chunk.read
+          w.flush
+        end
       end
     end
   end

data/lib/fluent/plugin/webhdfs_compressor_zstd.rb

@@ -0,0 +1,24 @@
+module Fluent::Plugin
+  class WebHDFSOutput < Output
+    class ZstdCompressor < Compressor
+      WebHDFSOutput.register_compressor('zstd', self)
+
+      def initialize(options = {})
+        begin
+          require "zstandard"
+        rescue LoadError
+          raise Fluent::ConfigError, "Install zstandard gem before use of zstd compressor"
+        end
+      end
+
+      def ext
+        ".zst"
+      end
+
+      def compress(chunk, tmp)
+        tmp.binmode
+        tmp.write Zstandard.deflate(chunk.read)
+      end
+    end
+  end
+end

data/test/plugin/test_out_webhdfs.rb

@@ -107,6 +107,7 @@ class WebHDFSOutputTest < Test::Unit::TestCase
     data(gzip: [:gzip, Fluent::Plugin::WebHDFSOutput::GzipCompressor],
          bzip2: [:bzip2, Fluent::Plugin::WebHDFSOutput::Bzip2Compressor],
          snappy: [:snappy, Fluent::Plugin::WebHDFSOutput::SnappyCompressor],
+         hadoop_snappy: [:hadoop_snappy, Fluent::Plugin::WebHDFSOutput::HadoopSnappyCompressor],
          lzo: [:lzo_command, Fluent::Plugin::WebHDFSOutput::LZOCommandCompressor])
     def test_compress(data)
       compress_type, compressor_class = data
@@ -126,6 +127,43 @@ class WebHDFSOutputTest < Test::Unit::TestCase
       assert_equal '/hdfs/path/file.%Y%m%d.%H%M.log', d.instance.path
       assert_equal compress_type, d.instance.compress
       assert_equal compressor_class, d.instance.compressor.class
+
+      time = event_time("2020-10-03 15:07:00 +0300")
+      metadata = d.instance.metadata("test", time, {})
+      chunk = d.instance.buffer.generate_chunk(metadata)
+      assert_equal "/hdfs/path/file.20201003.1507.log#{d.instance.compressor.ext}", d.instance.generate_path(chunk)
+    end
+
+    def test_explicit_extensions
+      conf = config_element(
+        "ROOT", "", {
+          "host" => "namenode.local",
+          "path" => "/hdfs/path/file.%Y%m%d.log",
+          "compress" => "snappy",
+          "extension" => ".snappy"
+        })
+      d = create_driver(conf)
+      time = event_time("2020-10-07 15:15:00 +0300")
+      metadata = d.instance.metadata("test", time, {})
+      chunk = d.instance.buffer.generate_chunk(metadata)
+      assert_equal "/hdfs/path/file.20201007.log.snappy", d.instance.generate_path(chunk)
+    end
+
+    data(snappy: [:snappy, Fluent::Plugin::WebHDFSOutput::SnappyCompressor],
+         hadoop_snappy: [:hadoop_snappy, Fluent::Plugin::WebHDFSOutput::HadoopSnappyCompressor])
+    def test_compression_block_size(data)
+      compress_type, compressor_class = data
+      conf = config_element(
+        "ROOT", "", {
+          "host" => "namenode.local",
+          "path" => "/hdfs/path/file.%Y%m%d.log",
+          "compress" => compress_type,
+          "block_size" => 16384
+        })
+      d = create_driver(conf)
+
+      assert_equal compress_type, d.instance.compress
+      assert_equal 16384, d.instance.compressor.block_size
     end
 
     def test_placeholders_old_style

data/test/plugin/test_snappy_compressors.rb

@@ -0,0 +1,72 @@
+require "helper"
+require "fluent/plugin/buf_memory"
+begin
+  require "snappy"
+rescue LoadError
+end
+
+class SnappyCompressorsTest < Test::Unit::TestCase
+  class Snappy < self
+
+    CONFIG = %[
+      host namenode.local
+      path /hdfs/path/file.%Y%m%d.log
+    ]
+
+    def setup
+      omit unless Object.const_defined?(:Snappy)
+      Fluent::Test.setup
+
+      @compressors_size = 2
+      @compressors = [
+        Fluent::Plugin::WebHDFSOutput::SnappyCompressor.new,
+        Fluent::Plugin::WebHDFSOutput::HadoopSnappyCompressor.new
+      ]
+      @readers = [
+        ::Snappy::Reader,
+        ::Snappy::Hadoop::Reader
+      ]
+      @exts = [".sz", ".snappy"]
+    end
+
+    def create_driver(conf = CONFIG)
+      Fluent::Test::Driver::Output.new(Fluent::Plugin::WebHDFSOutput).configure(conf)
+    end
+
+    def test_ext
+      for i in 0...@compressors_size do
+        assert_equal(@exts[i], @compressors[i].ext)
+      end
+    end
+
+    def test_compress
+      d = create_driver
+      if d.instance.respond_to?(:buffer)
+        buffer = d.instance.buffer
+      else
+        buffer = d.instance.instance_variable_get(:@buffer)
+      end
+
+      if buffer.respond_to?(:generate_chunk)
+        chunk = buffer.generate_chunk("test")
+        chunk.concat("hello snappy\n" * 32 * 1024, 1)
+      else
+        chunk = buffer.new_chunk("test")
+        chunk << "hello snappy\n" * 32 * 1024
+      end
+
+      for i in 0...@compressors_size do
+        io = Tempfile.new("snappy-")
+        @compressors[i].compress(chunk, io)
+        io.open
+        chunk_bytesize = chunk.respond_to?(:bytesize) ? chunk.bytesize : chunk.size
+        assert(chunk_bytesize > io.read.bytesize)
+        io.rewind
+        reader = @readers[i].new(io)
+        assert_equal(chunk.read, reader.read)
+        io.close
+      end
+    end
+  end
+end
+

data/test/plugin/{test_compressor.rb → test_zstd_compressor.rb}

@@ -1,12 +1,9 @@
 require "helper"
 require "fluent/plugin/buf_memory"
-begin
-  require "snappy"
-rescue LoadError
-end
+require 'zstandard'
 
-class CompressorTest < Test::Unit::TestCase
-  class Snappy < self
+class ZstdCompressorTest < Test::Unit::TestCase
+  class Zstd < self
 
     CONFIG = %[
       host namenode.local
@@ -14,9 +11,9 @@ class CompressorTest < Test::Unit::TestCase
     ]
 
     def setup
-      omit unless Object.const_defined?(:Snappy)
+      omit unless Object.const_defined?(:Zstandard)
       Fluent::Test.setup
-      @compressor = Fluent::Plugin::WebHDFSOutput::SnappyCompressor.new
+      @compressor = Fluent::Plugin::WebHDFSOutput::ZstdCompressor.new
     end
 
     def create_driver(conf = CONFIG)
@@ -24,7 +21,7 @@ class CompressorTest < Test::Unit::TestCase
     end
 
     def test_ext
-      assert_equal(".sz", @compressor.ext)
+      assert_equal(".zst", @compressor.ext)
     end
 
     def test_compress
@@ -37,22 +34,22 @@ class CompressorTest < Test::Unit::TestCase
 
       if buffer.respond_to?(:generate_chunk)
         chunk = buffer.generate_chunk("test")
-        chunk.concat("hello snappy\n" * 32 * 1024, 1)
+        chunk.concat("hello gzip\n" * 32 * 1024, 1)
       else
         chunk = buffer.new_chunk("test")
-        chunk << "hello snappy\n" * 32 * 1024
+        chunk << "hello gzip\n" * 32 * 1024
       end
 
-      io = Tempfile.new("snappy-")
+      io = Tempfile.new("zstd-")
       @compressor.compress(chunk, io)
-      io.open
+      assert !io.closed?
       chunk_bytesize = chunk.respond_to?(:bytesize) ? chunk.bytesize : chunk.size
       assert(chunk_bytesize > io.read.bytesize)
       io.rewind
-      reader = ::Snappy::Reader.new(io)
-      assert_equal(chunk.read, reader.read)
+      reader = Zstandard.inflate(io.read)
+      io.rewind
+      assert_equal(chunk.read, reader)
       io.close
     end
   end
 end
-

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: fluent-plugin-webhdfs
 version: !ruby/object:Gem::Version
-  version: 1.2.4
+  version: 1.4.0
 platform: ruby
 authors:
 - TAGOMORI Satoshi
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2019-07-09 00:00:00.000000000 Z
+date: 2020-12-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake
@@ -94,6 +94,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: zstandard
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: fluentd
   requirement: !ruby/object:Gem::Requirement
@@ -140,13 +154,16 @@ files:
 - lib/fluent/plugin/out_webhdfs.rb
 - lib/fluent/plugin/webhdfs_compressor_bzip2.rb
 - lib/fluent/plugin/webhdfs_compressor_gzip.rb
+- lib/fluent/plugin/webhdfs_compressor_hadoop_snappy.rb
 - lib/fluent/plugin/webhdfs_compressor_lzo_command.rb
 - lib/fluent/plugin/webhdfs_compressor_snappy.rb
 - lib/fluent/plugin/webhdfs_compressor_text.rb
+- lib/fluent/plugin/webhdfs_compressor_zstd.rb
 - test/helper.rb
-- test/plugin/test_compressor.rb
 - test/plugin/test_gzip_compressor.rb
 - test/plugin/test_out_webhdfs.rb
+- test/plugin/test_snappy_compressors.rb
+- test/plugin/test_zstd_compressor.rb
 homepage: https://github.com/fluent/fluent-plugin-webhdfs
 licenses:
 - Apache-2.0
@@ -172,6 +189,7 @@ specification_version: 4
 summary: Fluentd plugin to write data on HDFS over WebHDFS, with flexible formatting
 test_files:
 - test/helper.rb
-- test/plugin/test_compressor.rb
 - test/plugin/test_gzip_compressor.rb
 - test/plugin/test_out_webhdfs.rb
+- test/plugin/test_snappy_compressors.rb
+- test/plugin/test_zstd_compressor.rb