logstash-filter-collate 0.1.0

checksums.yaml

@@ -0,0 +1,15 @@
+---
+!binary "U0hBMQ==":
+  metadata.gz: !binary |-
+    YTA3MjQ5ZGZjYTM1NTdmZGFiNjViNzI1NDllMmI2YjM2YzE1NmQ2ZQ==
+  data.tar.gz: !binary |-
+    MmZiNzk5MmM4NmM3ZDMxOGMzYzBjZWE1MjQ1YjI2ZjQ1YzRkOTVlOA==
+SHA512:
+  metadata.gz: !binary |-
+    ODk2YzZlMGM4NjM0NWY1OTk3ZTMyNzY0OGE4OTBhMmQ3NTQzOTQ0NzJhYTgw
+    MjQ3YzBiZmJmODJiODliYjNkYmUzNmMzNTZiM2M1YzM4ZjQwOWFjZGJlMDM1
+    NDlhZDdkMDIwNmFkODY5YmZhNmExOThjMzI3YjBhMmUwMmQ2MGY=
+  data.tar.gz: !binary |-
+    MTRlYTFjYTYwNzg5MDZlNjE2ZGMzMDI2YWQwZTM2YTEzOGZiMDQ2NWQ0MWI1
+    NjY2ZWMxNzY4Mjg0OGVmNzJmNDdjNTJjOGY1ZmMyZTg1OTc3MTI3OTNhMmY4
+    MmNiNjFmMTE4MmZkMGFmNTNhNjBkMTUxMzQwMzI0ODA1OWFiNGQ=

data/.gitignore

@@ -0,0 +1,4 @@
+*.gem
+Gemfile.lock
+.bundle
+vendor

data/Gemfile

@@ -0,0 +1,3 @@
+source 'http://rubygems.org'
+gem 'rake'
+gem 'gem_publisher'

data/LICENSE

@@ -0,0 +1,13 @@
+Copyright (c) 2012-2014 Elasticsearch <http://www.elasticsearch.org>
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.

data/Rakefile

@@ -0,0 +1,6 @@
+@files=[]
+
+task :default do
+  system("rake -T")
+end
+

data/lib/logstash/filters/collate.rb

@@ -0,0 +1,116 @@
+# encoding: utf-8
+require "logstash/filters/base"
+require "logstash/namespace"
+
+
+# Collate events by time or count.
+#
+# The original goal of this filter was to merge the logs from different sources
+# by the time of log, for example, in real-time log collection, logs can be
+# collated by amount of 3000 logs or can be collated in 30 seconds.
+#
+# The config looks like this:
+# [source,ruby]
+#     filter {
+#       collate {
+#         size => 3000
+#         interval => "30s"
+#         order => "ascending"
+#       }
+#     }
+class LogStash::Filters::Collate < LogStash::Filters::Base
+
+  config_name "collate"
+  milestone 1
+
+  # How many logs should be collated.
+  config :count, :validate => :number, :default => 1000
+
+  # The `interval` is the time window which how long the logs should be collated. (default `1m`)
+  config :interval, :validate => :string, :default => "1m"
+
+  # The `order` collated events should appear in.
+  config :order, :validate => ["ascending", "descending"], :default => "ascending"
+
+  public
+  def register
+    require "thread"
+    require "rufus/scheduler"
+
+    @mutex = Mutex.new
+    @collatingDone = false
+    @collatingArray = Array.new
+    @scheduler = Rufus::Scheduler.start_new
+    @job = @scheduler.every @interval do
+      @logger.info("Scheduler Activated")
+      @mutex.synchronize{
+        collate
+      }
+    end
+  end # def register
+
+  public
+  def filter(event)
+    @logger.info("do collate filter")
+    if event == LogStash::SHUTDOWN
+      @job.trigger()
+      @job.unschedule()
+      @logger.info("collate filter thread shutdown.")
+      return
+    end
+
+    # if the event is collated, a "collated" tag will be marked, so for those uncollated event, cancel them first.
+    if event["tags"].nil? || !event.tags.include?("collated")
+      event.cancel
+    else
+      return
+    end
+
+    @mutex.synchronize{
+      @collatingArray.push(event.clone)
+
+      if (@collatingArray.length == @count)
+        collate
+      end
+
+      if (@collatingDone)
+        while collatedEvent = @collatingArray.pop
+          collatedEvent["tags"] = Array.new if collatedEvent["tags"].nil?
+          collatedEvent["tags"] << "collated"
+          filter_matched(collatedEvent)
+          yield collatedEvent
+        end # while @collatingArray.pop
+        # reset collatingDone flag
+        @collatingDone = false
+      end
+    }
+  end # def filter
+
+  private
+  def collate
+    if (@order == "ascending")
+      # call .to_i for now until https://github.com/elasticsearch/logstash/issues/2052 is fixed
+      @collatingArray.sort! { |eventA, eventB| eventB.timestamp.to_i <=> eventA.timestamp.to_i }
+    else 
+      @collatingArray.sort! { |eventA, eventB| eventA.timestamp.to_i <=> eventB.timestamp.to_i }
+    end
+    @collatingDone = true
+  end # def collate
+
+  # Flush any pending messages.
+  public
+  def flush(options = {})
+    events = []
+    if (@collatingDone)
+      @mutex.synchronize{
+        while collatedEvent = @collatingArray.pop
+          collatedEvent["tags"] << "collated"
+          events << collatedEvent
+        end # while @collatingArray.pop
+      }
+      # reset collatingDone flag.
+      @collatingDone = false
+    end
+    return events
+  end # def flush
+end #

data/logstash-filter-collate.gemspec

@@ -0,0 +1,27 @@
+Gem::Specification.new do |s|
+
+  s.name            = 'logstash-filter-collate'
+  s.version         = '0.1.0'
+  s.licenses        = ['Apache License (2.0)']
+  s.summary         = "Collate events by time or count"
+  s.description     = "The original goal of this filter was to merge the logs from different sources by the time of log."
+  s.authors         = ["Elasticsearch"]
+  s.email           = 'richard.pijnenburg@elasticsearch.com'
+  s.homepage        = "http://logstash.net/"
+  s.require_paths = ["lib"]
+
+  # Files
+  s.files = `git ls-files`.split($\)+::Dir.glob('vendor/*')
+
+  # Tests
+  s.test_files = s.files.grep(%r{^(test|spec|features)/})
+
+  # Special flag to let us know this is actually a logstash plugin
+  s.metadata = { "logstash_plugin" => "true", "group" => "filter" }
+
+  # Gem dependencies
+  s.add_runtime_dependency 'logstash', '>= 1.4.0', '< 2.0.0'
+  s.add_runtime_dependency 'rufus-scheduler', ['~> 2.0.24']
+
+end
+

data/rakelib/publish.rake

@@ -0,0 +1,9 @@
+require "gem_publisher"
+
+desc "Publish gem to RubyGems.org"
+task :publish_gem do |t|
+  gem_file = Dir.glob(File.expand_path('../*.gemspec',File.dirname(__FILE__))).first
+  gem = GemPublisher.publish_if_updated(gem_file, :rubygems)
+  puts "Published #{gem}" if gem
+end
+

data/rakelib/vendor.rake

@@ -0,0 +1,169 @@
+require "net/http"
+require "uri"
+require "digest/sha1"
+
+def vendor(*args)
+  return File.join("vendor", *args)
+end
+
+directory "vendor/" => ["vendor"] do |task, args|
+  mkdir task.name
+end
+
+def fetch(url, sha1, output)
+
+  puts "Downloading #{url}"
+  actual_sha1 = download(url, output)
+
+  if actual_sha1 != sha1
+    fail "SHA1 does not match (expected '#{sha1}' but got '#{actual_sha1}')"
+  end
+end # def fetch
+
+def file_fetch(url, sha1)
+  filename = File.basename( URI(url).path )
+  output = "vendor/#{filename}"
+  task output => [ "vendor/" ] do
+    begin
+      actual_sha1 = file_sha1(output)
+      if actual_sha1 != sha1
+        fetch(url, sha1, output)
+      end
+    rescue Errno::ENOENT
+      fetch(url, sha1, output)
+    end
+  end.invoke
+
+  return output
+end
+
+def file_sha1(path)
+  digest = Digest::SHA1.new
+  fd = File.new(path, "r")
+  while true
+    begin
+      digest << fd.sysread(16384)
+    rescue EOFError
+      break
+    end
+  end
+  return digest.hexdigest
+ensure
+  fd.close if fd
+end
+
+def download(url, output)
+  uri = URI(url)
+  digest = Digest::SHA1.new
+  tmp = "#{output}.tmp"
+  Net::HTTP.start(uri.host, uri.port, :use_ssl => (uri.scheme == "https")) do |http|
+    request = Net::HTTP::Get.new(uri.path)
+    http.request(request) do |response|
+      fail "HTTP fetch failed for #{url}. #{response}" if [200, 301].include?(response.code)
+      size = (response["content-length"].to_i || -1).to_f
+      count = 0
+      File.open(tmp, "w") do |fd|
+        response.read_body do |chunk|
+          fd.write(chunk)
+          digest << chunk
+          if size > 0 && $stdout.tty?
+            count += chunk.bytesize
+            $stdout.write(sprintf("\r%0.2f%%", count/size * 100))
+          end
+        end
+      end
+      $stdout.write("\r      \r") if $stdout.tty?
+    end
+  end
+
+  File.rename(tmp, output)
+
+  return digest.hexdigest
+rescue SocketError => e
+  puts "Failure while downloading #{url}: #{e}"
+  raise
+ensure
+  File.unlink(tmp) if File.exist?(tmp)
+end # def download
+
+def untar(tarball, &block)
+  require "archive/tar/minitar"
+  tgz = Zlib::GzipReader.new(File.open(tarball))
+  # Pull out typesdb
+  tar = Archive::Tar::Minitar::Input.open(tgz)
+  tar.each do |entry|
+    path = block.call(entry)
+    next if path.nil?
+    parent = File.dirname(path)
+    
+    mkdir_p parent unless File.directory?(parent)
+
+    # Skip this file if the output file is the same size
+    if entry.directory?
+      mkdir path unless File.directory?(path)
+    else
+      entry_mode = entry.instance_eval { @mode } & 0777
+      if File.exists?(path)
+        stat = File.stat(path)
+        # TODO(sissel): Submit a patch to archive-tar-minitar upstream to
+        # expose headers in the entry.
+        entry_size = entry.instance_eval { @size }
+        # If file sizes are same, skip writing.
+        next if stat.size == entry_size && (stat.mode & 0777) == entry_mode
+      end
+      puts "Extracting #{entry.full_name} from #{tarball} #{entry_mode.to_s(8)}"
+      File.open(path, "w") do |fd|
+        # eof? check lets us skip empty files. Necessary because the API provided by
+        # Archive::Tar::Minitar::Reader::EntryStream only mostly acts like an
+        # IO object. Something about empty files in this EntryStream causes
+        # IO.copy_stream to throw "can't convert nil into String" on JRuby
+        # TODO(sissel): File a bug about this.
+        while !entry.eof?
+          chunk = entry.read(16384)
+          fd.write(chunk)
+        end
+          #IO.copy_stream(entry, fd)
+      end
+      File.chmod(entry_mode, path)
+    end
+  end
+  tar.close
+  File.unlink(tarball) if File.file?(tarball)
+end # def untar
+
+def ungz(file)
+
+  outpath = file.gsub('.gz', '')
+  tgz = Zlib::GzipReader.new(File.open(file))
+  begin
+    File.open(outpath, "w") do |out|
+      IO::copy_stream(tgz, out)
+    end
+    File.unlink(file)
+  rescue
+    File.unlink(outpath) if File.file?(outpath)
+   raise
+  end
+  tgz.close
+end
+
+desc "Process any vendor files required for this plugin"
+task "vendor" do |task, args|
+
+  @files.each do |file| 
+    download = file_fetch(file['url'], file['sha1'])
+    if download =~ /.tar.gz/
+      prefix = download.gsub('.tar.gz', '').gsub('vendor/', '')
+      untar(download) do |entry|
+        if !file['files'].nil?
+          next unless file['files'].include?(entry.full_name.gsub(prefix, ''))
+          out = entry.full_name.split("/").last
+        end
+        File.join('vendor', out)
+      end
+    elsif download =~ /.gz/
+      ungz(download)
+    end
+  end
+
+end

data/spec/filters/collate_spec.rb

@@ -0,0 +1,122 @@
+require "spec_helper"
+require "logstash/filters/collate"
+
+describe LogStash::Filters::Collate do
+  
+
+  describe "collate when count is full" do
+    config <<-CONFIG
+      filter {
+        collate {
+          count => 2
+        }
+      }
+    CONFIG
+
+    events = [
+      {
+        "@timestamp" => Time.iso8601("2013-01-02T00:00:00.000Z"),
+        "message" => "later message"
+      },
+      {
+        "@timestamp" => Time.iso8601("2013-01-01T00:00:00.000Z"),
+        "message" => "earlier message"
+      }
+    ]
+
+    sample(events) do
+      insist { subject }.is_a? Array
+      insist { subject.length } == 2
+      subject.each_with_index do |s,i|
+        if i == 0 # first one should be the earlier message
+          insist { s["message"] } == "earlier message"
+        end
+        if i == 1 # second one should be the later message
+          insist { s["message"]} == "later message"
+        end
+      end
+    end
+  end
+
+  describe "collate by desc" do
+    config <<-CONFIG
+      filter {
+        collate {
+          count => 3
+          order => "descending"
+        }
+      }
+    CONFIG
+
+    events = [
+      {
+        "@timestamp" => Time.iso8601("2013-01-03T00:00:00.000Z"),
+        "message" => "third message"
+      },
+      {
+        "@timestamp" => Time.iso8601("2013-01-01T00:00:00.000Z"),
+        "message" => "first message"
+      },
+      {
+        "@timestamp" => Time.iso8601("2013-01-02T00:00:00.000Z"),
+        "message" => "second message"
+      }
+    ]
+
+    sample(events) do
+      insist { subject }.is_a? Array
+      insist { subject.length } == 3
+      subject.each_with_index do |s,i|
+        if i == 0 # first one should be the third message
+          insist { s["message"] } == "third message"
+        end
+        if i == 1 # second one should be the second message
+          insist { s["message"]} == "second message"
+        end
+        if i == 2 # third one should be the third message
+          insist { s["message"]} == "first message"
+        end
+      end
+    end
+  end
+
+  # (Ignored) Currently this case can't pass because of the case depends on the flush function of the filter in the test, 
+  # there was a TODO marked in the code (spec_helper.rb, # TODO(sissel): pipeline flush needs to be implemented.), 
+  # and the case wants to test the scenario which collate was triggered by a scheduler, so in this case, it needs to sleep few seconds 
+  # waiting the scheduler triggered, and after the events were flushed, then the result can be checked.
+
+  # describe "collate when interval reached" do
+  #   config <<-CONFIG
+  #     filter {
+  #       collate {
+  #         interval => "1s"
+  #       }
+  #     }
+  #   CONFIG
+
+  #   events = [
+  #     {
+  #       "@timestamp" => Time.iso8601("2013-01-02T00:00:00.000Z"),
+  #       "message" => "later message"
+  #     },
+  #     {
+  #       "@timestamp" => Time.iso8601("2013-01-01T00:00:00.000Z"),
+  #       "message" => "earlier message"
+  #     }
+  #   ]
+
+  #   sample(events) do
+  #     sleep(2)
+  #     insist { subject }.is_a? Array
+  #     insist { subject.length } == 2
+  #     subject.each_with_index do |s,i|
+  #       if i == 0 # first one should be the earlier message
+  #         insist { s["message"] } == "earlier message"
+  #       end
+  #       if i == 1 # second one should be the later message
+  #         insist { s["message"]} == "later message"
+  #       end
+  #     end
+  #   end
+  # end
+end

metadata

@@ -0,0 +1,90 @@
+--- !ruby/object:Gem::Specification
+name: logstash-filter-collate
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- Elasticsearch
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2014-11-11 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: logstash
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 1.4.0
+    - - <
+      - !ruby/object:Gem::Version
+        version: 2.0.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 1.4.0
+    - - <
+      - !ruby/object:Gem::Version
+        version: 2.0.0
+- !ruby/object:Gem::Dependency
+  name: rufus-scheduler
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 2.0.24
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 2.0.24
+description: The original goal of this filter was to merge the logs from different
+  sources by the time of log.
+email: richard.pijnenburg@elasticsearch.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- .gitignore
+- Gemfile
+- LICENSE
+- Rakefile
+- lib/logstash/filters/collate.rb
+- logstash-filter-collate.gemspec
+- rakelib/publish.rake
+- rakelib/vendor.rake
+- spec/filters/collate_spec.rb
+homepage: http://logstash.net/
+licenses:
+- Apache License (2.0)
+metadata:
+  logstash_plugin: 'true'
+  group: filter
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.4.1
+signing_key: 
+specification_version: 4
+summary: Collate events by time or count
+test_files:
+- spec/filters/collate_spec.rb