logstash-input-elasticsearch 0.1.0

checksums.yaml

@@ -0,0 +1,15 @@
+---
+!binary "U0hBMQ==":
+  metadata.gz: !binary |-
+    ZGM3OGYyOWE4NTc1ODgxNGFmOTBkNzI1MWExYmMyOWExZTZlZjQyNg==
+  data.tar.gz: !binary |-
+    YjczYjE1ZWRhMjg4NmMzNjg0M2EwZTExMjEyNDI4YTk4MzU1MmZhYw==
+SHA512:
+  metadata.gz: !binary |-
+    NDNiNmFhMDRjMjEwMjAyMDg3NmE3ZmM2M2I4MzdiMThhMDI0Mjk0YmJiYWY1
+    NzFjNjM1YTcyOWYyY2NkMzkwMTVhOGQ5N2NmNTZjZGVmODQ3YTYwN2VmMDAy
+    MGM5NTViNThhNzJmNWU4NWYyOTg0MTk2NjQwOWFkYzY5Zjk2YTM=
+  data.tar.gz: !binary |-
+    NGYxYjNlY2I4NjJiNWU1YTkyZjYwYmNjYzFhNGE3ZDI1MWY3N2VlMTFjNjBm
+    OGEzNWUyZjQ2Yjc0ZmJmNGU0NTVmYTE2Y2JhMTlhMDgwY2UzZjBhZWFkZTI4
+    ZmYwNTk4MmNiMjRhYWVlMzEyNzEyMDI0MjYyMTM3OTQwZDY3MWE=

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/inputs/elasticsearch.rb

@@ -0,0 +1,134 @@
+# encoding: utf-8
+require "logstash/inputs/base"
+require "logstash/namespace"
+require "logstash/util/socket_peer"
+require "logstash/json"
+
+# Read from an Elasticsearch cluster, based on search query results.
+# This is useful for replaying test logs, reindexing, etc.
+#
+# Example:
+#
+#     input {
+#       # Read all documents from Elasticsearch matching the given query
+#       elasticsearch {
+#         host => "localhost"
+#         query => "ERROR"
+#       }
+#     }
+#
+# This would create an Elasticsearch query with the following format:
+#
+#     http://localhost:9200/logstash-*/_search?q=ERROR&scroll=1m&size=1000
+#
+# * TODO(sissel): Option to keep the index, type, and doc id so we can do reindexing?
+class LogStash::Inputs::Elasticsearch < LogStash::Inputs::Base
+  config_name "elasticsearch"
+  milestone 1
+
+  default :codec, "json"
+
+  # The IP address or hostname of your Elasticsearch server.
+  config :host, :validate => :string, :required => true
+
+  # The HTTP port of your Elasticsearch server's REST interface.
+  config :port, :validate => :number, :default => 9200
+
+  # The index or alias to search.
+  config :index, :validate => :string, :default => "logstash-*"
+
+  # The query to be executed.
+  config :query, :validate => :string, :default => "*"
+
+  # Enable the Elasticsearch "scan" search type.  This will disable
+  # sorting but increase speed and performance.
+  config :scan, :validate => :boolean, :default => true
+
+  # This allows you to set the maximum number of hits returned per scroll.
+  config :size, :validate => :number, :default => 1000
+
+  # This parameter controls the keepalive time in seconds of the scrolling
+  # request and initiates the scrolling process. The timeout applies per
+  # round trip (i.e. between the previous scan scroll request, to the next).
+  config :scroll, :validate => :string, :default => "1m"
+
+  public
+  def register
+    require "ftw"
+    @agent = FTW::Agent.new
+
+    params = {
+      "q" => @query,
+      "scroll" => @scroll,
+      "size" => "#{@size}",
+    }
+    params['search_type'] = "scan" if @scan
+
+    @search_url = "http://#{@host}:#{@port}/#{@index}/_search?#{encode(params)}"
+    @scroll_url = "http://#{@host}:#{@port}/_search/scroll?#{encode({"scroll" => @scroll})}"
+  end # def register
+
+  private
+  def encode(hash)
+    return hash.collect do |key, value|
+      CGI.escape(key) + "=" + CGI.escape(value)
+    end.join("&")
+  end # def encode
+
+  private
+  def execute_search_request
+    response = @agent.get!(@search_url)
+    json = ""
+    response.read_body { |c| json << c }
+    json
+  end
+
+  private
+  def execute_scroll_request(scroll_id)
+    response = @agent.post!(@scroll_url, :body => scroll_id)
+    json = ""
+    response.read_body { |c| json << c }
+    json
+  end
+
+  public
+  def run(output_queue)
+    result = LogStash::Json.load(execute_search_request)
+    scroll_id = result["_scroll_id"]
+
+    # When using the search_type=scan we don't get an initial result set.
+    # So we do it here.
+    if @scan
+      result = LogStash::Json.load(execute_scroll_request(scroll_id))
+    end
+
+    loop do
+      break if result.nil?
+      hits = result["hits"]["hits"]
+      break if hits.empty?
+
+      hits.each do |hit|
+        # Hack to make codecs work
+        @codec.decode(LogStash::Json.dump(hit["_source"])) do |event|
+          decorate(event)
+          output_queue << event
+        end
+      end
+
+      # Get the scroll id from the previous result set and use it for getting the next data set
+      scroll_id = result["_scroll_id"]
+
+      # Fetch the next result set
+      result = LogStash::Json.load(execute_scroll_request(scroll_id))
+
+      if result["error"]
+        @logger.warn(result["error"], :request => scroll_url)
+        # TODO(sissel): raise an error instead of breaking
+        break
+      end
+
+    end
+  rescue LogStash::ShutdownSignal
+    # Do nothing, let us quit.
+  end # def run
+end # class LogStash::Inputs::Elasticsearch

data/logstash-input-elasticsearch.gemspec

@@ -0,0 +1,29 @@
+Gem::Specification.new do |s|
+
+  s.name            = 'logstash-input-elasticsearch'
+  s.version         = '0.1.0'
+  s.licenses        = ['Apache License (2.0)']
+  s.summary         = "Read from an Elasticsearch cluster, based on search query results"
+  s.description     = "Read from an Elasticsearch cluster, based on search query results"
+  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" => "input" }
+
+  # Gem dependencies
+  s.add_runtime_dependency 'logstash', '>= 1.4.0', '< 2.0.0'
+
+  s.add_runtime_dependency 'ftw', ['~> 0.0.39']
+  s.add_runtime_dependency 'logstash-codec-json'
+
+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/inputs/elasticsearch_spec.rb

@@ -0,0 +1,80 @@
+require "spec_helper"
+require "logstash/inputs/elasticsearch"
+
+describe "inputs/elasticsearch" do
+  
+
+  search_response = <<-RESPONSE
+    {
+      "_scroll_id":"xxx",
+      "took":5,
+      "timed_out":false,
+      "_shards":{"total":15,"successful":15,"failed":0},
+      "hits":{
+        "total":1000050,
+        "max_score":1.0,
+        "hits":[
+          {
+            "_index":"logstash2",
+            "_type":"logs",
+            "_id":"AmaqL7VuSWKF-F6N_Gz72g",
+            "_score":1.0,
+            "_source" : {
+              "message":"foobar",
+              "@version":"1",
+              "@timestamp":"2014-05-19T21:08:39.000Z",
+              "host":"colin-mbp13r"
+            }
+          }
+        ]
+      }
+    }
+  RESPONSE
+
+  scroll_response = <<-RESPONSE
+    {
+      "hits":{
+        "hits":[]
+      }
+    }
+  RESPONSE
+
+  config <<-CONFIG
+    input {
+      elasticsearch {
+        host => "localhost"
+        scan => false
+      }
+    }
+  CONFIG
+
+  it "should retrieve json event from elasticseach" do
+    # I somewhat duplicated our "input" rspec extension because I needed to add mocks for the the actual ES calls
+    # and rspec expectations need to be in "it" statement but the "input" extension defines the "it"
+    # TODO(colin) see how we can improve our rspec extension to better integrate in these scenarios
+
+    expect_any_instance_of(LogStash::Inputs::Elasticsearch).to receive(:execute_search_request).and_return(search_response)
+    expect_any_instance_of(LogStash::Inputs::Elasticsearch).to receive(:execute_scroll_request).with(any_args).and_return(scroll_response)
+
+    pipeline = LogStash::Pipeline.new(config)
+    queue = Queue.new
+    pipeline.instance_eval do
+      @output_func = lambda { |event| queue << event }
+    end
+    pipeline_thread = Thread.new { pipeline.run }
+    event = queue.pop
+
+    insist { event["message"] } == "foobar"
+
+    # do not call pipeline.shutdown here, as it will stop the plugin execution randomly
+    # and maybe kill input before calling execute_scroll_request.
+    # TODO(colin) we should rework the pipeliene shutdown to allow a soft/clean shutdown mecanism,
+    # using a shutdown event which can be fed into each plugin queue and when the plugin sees it
+    # exits after completing its processing.
+    #
+    # pipeline.shutdown
+    #
+    # instead, since our scroll_response will terminate the plugin, we can just join the pipeline thread
+    pipeline_thread.join
+  end
+end

metadata

@@ -0,0 +1,102 @@
+--- !ruby/object:Gem::Specification
+name: logstash-input-elasticsearch
+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
+- !ruby/object:Gem::Dependency
+  name: ftw
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.0.39
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.0.39
+- !ruby/object:Gem::Dependency
+  name: logstash-codec-json
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+description: Read from an Elasticsearch cluster, based on search query results
+email: richard.pijnenburg@elasticsearch.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- .gitignore
+- Gemfile
+- Rakefile
+- lib/logstash/inputs/elasticsearch.rb
+- logstash-input-elasticsearch.gemspec
+- rakelib/publish.rake
+- rakelib/vendor.rake
+- spec/inputs/elasticsearch_spec.rb
+homepage: http://logstash.net/
+licenses:
+- Apache License (2.0)
+metadata:
+  logstash_plugin: 'true'
+  group: input
+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: Read from an Elasticsearch cluster, based on search query results
+test_files:
+- spec/inputs/elasticsearch_spec.rb