logstash-output-graphite 0.1.0

checksums.yaml

@@ -0,0 +1,15 @@
+---
+!binary "U0hBMQ==":
+  metadata.gz: !binary |-
+    ZTM5YjJhNGYyNTBkZjZiZmNmZTAyNWZiMWRmNzRjODRlYjZhNmE4OA==
+  data.tar.gz: !binary |-
+    NTI0YjBlMDY4NmUyNDY3ODEyNDc1ZjM2NzhmMDVmNmRkMzhiZDliYQ==
+SHA512:
+  metadata.gz: !binary |-
+    ODU3OGUzNDllOGNiMjc5NGRjMDQyNDYyNzlkYTFmOGExNDE2YTkwN2YxZDQ5
+    M2MyYjBhMzQ5OThjZTkyMzllNTJiZGEzOTI4ODAzMzZkMjJmNzBlMmJkYzc1
+    NDBhODI5ZGNmNTAyZGJmYTkwZThhYWU3Y2M2YTVlNTQ1N2YyMjQ=
+  data.tar.gz: !binary |-
+    NmUzZWFkZTg5N2M4NDRmNjMwNjBlNmM2OWYzODRmM2Y5N2FlODYxM2RmNGU3
+    M2IzOTliYTY5YmIwOGRkMGU5OWViMDUzMTlkMTYyMGU4NGI0NTNkZTI2NTRl
+    ZDJlMTVmOGJhNzQ4MzE1YTc2Mzc3OWE5ZWQyYmMwOGUwZDdhNjM=

data/.gitignore

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

data/Gemfile

@@ -0,0 +1,4 @@
+source 'http://rubygems.org'
+gem 'rake'
+gem 'gem_publisher'
+gem 'archive-tar-minitar'

data/Rakefile

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

data/lib/logstash/outputs/graphite.rb

@@ -0,0 +1,146 @@
+# encoding: utf-8
+require "logstash/outputs/base"
+require "logstash/namespace"
+require "socket"
+
+# This output allows you to pull metrics from your logs and ship them to
+# Graphite. Graphite is an open source tool for storing and graphing metrics.
+#
+# An example use case: Some applications emit aggregated stats in the logs
+# every 10 seconds. Using the grok filter and this output, it is possible to
+# capture the metric values from the logs and emit them to Graphite.
+class LogStash::Outputs::Graphite < LogStash::Outputs::Base
+  config_name "graphite"
+  milestone 2
+
+  EXCLUDE_ALWAYS = [ "@timestamp", "@version" ]
+
+  DEFAULT_METRICS_FORMAT = "*"
+  METRIC_PLACEHOLDER = "*"
+
+  # The hostname or IP address of the Graphite server.
+  config :host, :validate => :string, :default => "localhost"
+
+  # The port to connect to on the Graphite server.
+  config :port, :validate => :number, :default => 2003
+
+  # Interval between reconnect attempts to Carbon.
+  config :reconnect_interval, :validate => :number, :default => 2
+
+  # Should metrics be resent on failure?
+  config :resend_on_failure, :validate => :boolean, :default => false
+
+  # The metric(s) to use. This supports dynamic strings like %{host}
+  # for metric names and also for values. This is a hash field with key 
+  # being the metric name, value being the metric value. Example:
+  #
+  #     [ "%{host}/uptime", "%{uptime_1m}" ]
+  #
+  # The value will be coerced to a floating point value. Values which cannot be
+  # coerced will be set to zero (0). You may use either `metrics` or `fields_are_metrics`,
+  # but not both.
+  config :metrics, :validate => :hash, :default => {}
+
+  # An array indicating that these event fields should be treated as metrics
+  # and will be sent verbatim to Graphite. You may use either `fields_are_metrics`
+  # or `metrics`, but not both.
+  config :fields_are_metrics, :validate => :boolean, :default => false
+
+  # Include only regex matched metric names.
+  config :include_metrics, :validate => :array, :default => [ ".*" ]
+
+  # Exclude regex matched metric names, by default exclude unresolved %{field} strings.
+  config :exclude_metrics, :validate => :array, :default => [ "%\{[^}]+\}" ]
+
+  # Enable debug output.
+  config :debug, :validate => :boolean, :default => false, :deprecated => "This setting was never used by this plugin. It will be removed soon."
+
+  # Defines the format of the metric string. The placeholder '*' will be
+  # replaced with the name of the actual metric.
+  #
+  #     metrics_format => "foo.bar.*.sum"
+  #
+  # NOTE: If no metrics_format is defined, the name of the metric will be used as fallback.
+  config :metrics_format, :validate => :string, :default => DEFAULT_METRICS_FORMAT
+
+  def register
+    @include_metrics.collect!{|regexp| Regexp.new(regexp)}
+    @exclude_metrics.collect!{|regexp| Regexp.new(regexp)}
+
+    if @metrics_format && !@metrics_format.include?(METRIC_PLACEHOLDER)
+      @logger.warn("metrics_format does not include placeholder #{METRIC_PLACEHOLDER} .. falling back to default format: #{DEFAULT_METRICS_FORMAT.inspect}")
+
+      @metrics_format = DEFAULT_METRICS_FORMAT
+    end
+
+    connect
+  end # def register
+
+  def connect
+    # TODO(sissel): Test error cases. Catch exceptions. Find fortune and glory. Retire to yak farm.
+    begin
+      @socket = TCPSocket.new(@host, @port)
+    rescue Errno::ECONNREFUSED => e
+      @logger.warn("Connection refused to graphite server, sleeping...",
+                   :host => @host, :port => @port)
+      sleep(@reconnect_interval)
+      retry
+    end
+  end # def connect
+
+  def construct_metric_name(metric)
+    if @metrics_format
+      return @metrics_format.gsub(METRIC_PLACEHOLDER, metric)
+    end
+
+    metric
+  end
+
+  public
+  def receive(event)
+    return unless output?(event)
+
+    # Graphite message format: metric value timestamp\n
+
+    messages = []
+    timestamp = event.sprintf("%{+%s}")
+
+    if @fields_are_metrics
+      @logger.debug("got metrics event", :metrics => event.to_hash)
+      event.to_hash.each do |metric,value|
+        next if EXCLUDE_ALWAYS.include?(metric)
+        next unless @include_metrics.empty? || @include_metrics.any? { |regexp| metric.match(regexp) }
+        next if @exclude_metrics.any? {|regexp| metric.match(regexp)}
+        messages << "#{construct_metric_name(metric)} #{event.sprintf(value.to_s).to_f} #{timestamp}"
+      end
+    else
+      @metrics.each do |metric, value|
+        @logger.debug("processing", :metric => metric, :value => value)
+        metric = event.sprintf(metric)
+        next unless @include_metrics.any? {|regexp| metric.match(regexp)}
+        next if @exclude_metrics.any? {|regexp| metric.match(regexp)}
+        messages << "#{construct_metric_name(event.sprintf(metric))} #{event.sprintf(value).to_f} #{timestamp}"
+      end
+    end
+
+    if messages.empty?
+      @logger.debug("Message is empty, not sending anything to Graphite", :messages => messages, :host => @host, :port => @port)
+    else
+      message = messages.join("\n")
+      @logger.debug("Sending carbon messages", :messages => messages, :host => @host, :port => @port)
+
+      # Catch exceptions like ECONNRESET and friends, reconnect on failure.
+      # TODO(sissel): Test error cases. Catch exceptions. Find fortune and glory.
+      begin
+        @socket.puts(message)
+      rescue Errno::EPIPE, Errno::ECONNRESET => e
+        @logger.warn("Connection to graphite server died",
+                     :exception => e, :host => @host, :port => @port)
+        sleep(@reconnect_interval)
+        connect
+        retry if @resend_on_failure
+      end
+    end
+
+  end # def receive
+end # class LogStash::Outputs::Graphite

data/logstash-output-graphite.gemspec

@@ -0,0 +1,26 @@
+Gem::Specification.new do |s|
+
+  s.name            = 'logstash-output-graphite'
+  s.version         = '0.1.0'
+  s.licenses        = ['Apache License (2.0)']
+  s.summary         = "This output allows you to pull metrics from your logs and ship them to Graphite"
+  s.description     = "This output allows you to pull metrics from your logs and ship them to Graphite"
+  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" => "output" }
+
+  # Gem dependencies
+  s.add_runtime_dependency 'logstash', '>= 1.4.0', '< 2.0.0'
+
+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/outputs/graphite_spec.rb

@@ -0,0 +1,236 @@
+require "spec_helper"
+require "logstash/outputs/graphite"
+require "mocha/api"
+
+describe LogStash::Outputs::Graphite, :socket => true do
+  
+
+  describe "defaults should include all metrics" do
+    port = 4939
+    config <<-CONFIG
+      input {
+        generator {
+          message => "foo=fancy bar=42"
+          count => 1
+          type => "generator"
+        }
+      }
+
+      filter {
+        kv { }
+      }
+
+      output {
+        graphite {
+          host => "localhost"
+          port => #{port}
+          metrics => [ "hurray.%{foo}", "%{bar}" ]
+        }
+      }
+    CONFIG
+
+    let(:queue) { Queue.new }
+    before :each do
+      server = TCPServer.new("127.0.0.1", port)
+      Thread.new do
+        client = server.accept
+        p client
+        while true
+          p :read
+          line = client.readline
+          p :done
+          queue << line
+          p line
+        end
+      end
+    end
+
+    agent do
+      lines = queue.pop
+
+      insist { lines.size } == 1
+      insist { lines }.any? { |l| l =~ /^hurray.fancy 42.0 \d{10,}\n$/ }
+    end
+  end
+
+  describe "fields_are_metrics => true" do
+    describe "metrics_format => ..." do
+      describe "match one key" do
+        config <<-CONFIG
+          input {
+            generator {
+              message => "foo=123"
+              count => 1
+              type => "generator"
+            }
+          }
+
+          filter {
+            kv { }
+          }
+
+          output {
+            graphite {
+                host => "localhost"
+                port => 2003
+                fields_are_metrics => true
+                include_metrics => ["foo"]
+                metrics_format => "foo.bar.sys.data.*"
+                debug => true
+            }
+          }
+        CONFIG
+
+        agent do
+          @mock.rewind
+          lines = @mock.readlines
+          insist { lines.size } == 1
+          insist { lines[0] } =~ /^foo.bar.sys.data.foo 123.0 \d{10,}\n$/
+        end
+      end
+
+      describe "match all keys" do
+        config <<-CONFIG
+          input {
+            generator {
+              message => "foo=123 bar=42"
+              count => 1
+              type => "generator"
+            }
+          }
+
+          filter {
+            kv { }
+          }
+
+          output {
+            graphite {
+                host => "localhost"
+                port => 2003
+                fields_are_metrics => true
+                include_metrics => [".*"]
+                metrics_format => "foo.bar.sys.data.*"
+                debug => true
+            }
+          }
+        CONFIG
+
+        agent do
+          @mock.rewind
+          lines = @mock.readlines.delete_if { |l| l =~ /\.sequence \d+/ }
+
+          insist { lines.size } == 2
+          insist { lines }.any? { |l| l =~ /^foo.bar.sys.data.foo 123.0 \d{10,}\n$/ }
+          insist { lines }.any? { |l| l =~ /^foo.bar.sys.data.bar 42.0 \d{10,}\n$/ }
+        end
+      end
+
+      describe "no match" do
+        config <<-CONFIG
+          input {
+            generator {
+              message => "foo=123 bar=42"
+              count => 1
+              type => "generator"
+            }
+          }
+
+          filter {
+            kv { }
+          }
+
+          output {
+            graphite {
+              host => "localhost"
+              port => 2003
+              fields_are_metrics => true
+              include_metrics => ["notmatchinganything"]
+              metrics_format => "foo.bar.sys.data.*"
+              debug => true
+            }
+          }
+        CONFIG
+
+        agent do
+          @mock.rewind
+          lines = @mock.readlines
+          insist { lines.size } == 0
+        end
+      end
+
+      describe "match one key with invalid metric_format" do
+        config <<-CONFIG
+          input {
+            generator {
+              message => "foo=123"
+              count => 1
+              type => "generator"
+            }
+          }
+
+          filter {
+            kv { }
+          }
+
+          output {
+            graphite {
+                host => "localhost"
+                port => 2003
+                fields_are_metrics => true
+                include_metrics => ["foo"]
+                metrics_format => "invalidformat"
+                debug => true
+            }
+          }
+        CONFIG
+
+        agent do
+          @mock.rewind
+          lines = @mock.readlines
+          insist { lines.size } == 1
+          insist { lines[0] } =~ /^foo 123.0 \d{10,}\n$/
+        end
+      end
+    end
+  end
+
+  describe "fields are metrics = false" do
+    describe "metrics_format not set" do
+      describe "match one key with metrics list" do
+        config <<-CONFIG
+          input {
+            generator {
+              message => "foo=123"
+              count => 1
+              type => "generator"
+            }
+          }
+
+          filter {
+            kv { }
+          }
+
+          output {
+            graphite {
+                host => "localhost"
+                port => 2003
+                fields_are_metrics => false
+                include_metrics => ["foo"]
+                metrics => [ "custom.foo", "%{foo}" ]
+                debug => true
+            }
+          }
+        CONFIG
+
+        agent do
+          @mock.rewind
+          lines = @mock.readlines
+
+          insist { lines.size } == 1
+          insist { lines[0] } =~ /^custom.foo 123.0 \d{10,}\n$/
+        end
+      end
+
+    end
+  end
+end

metadata

@@ -0,0 +1,75 @@
+--- !ruby/object:Gem::Specification
+name: logstash-output-graphite
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- Elasticsearch
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2014-11-03 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
+description: This output allows you to pull metrics from your logs and ship them to
+  Graphite
+email: richard.pijnenburg@elasticsearch.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- .gitignore
+- Gemfile
+- Rakefile
+- lib/logstash/outputs/graphite.rb
+- logstash-output-graphite.gemspec
+- rakelib/publish.rake
+- rakelib/vendor.rake
+- spec/outputs/graphite_spec.rb
+homepage: http://logstash.net/
+licenses:
+- Apache License (2.0)
+metadata:
+  logstash_plugin: 'true'
+  group: output
+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: This output allows you to pull metrics from your logs and ship them to Graphite
+test_files:
+- spec/outputs/graphite_spec.rb