logstash-output-googlecloudstorage 2.1.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: abbcf0973e9998e05af85da8a3b5a099b59301e85e8a333f797962d9578b5d4b
+  data.tar.gz: 432e21ad4d816480a1b3c16da6c75f265742de83e5d647b791a9be7c189be42e
+SHA512:
+  metadata.gz: 1474e3e8f52436f67319f2c4ba99d6ad65f7dbf22d44cf5a080a809d0e3935240cf1b5297b6cbc61a8061ce0e4a86d1848c9358e195c825b56da8334c05f8459
+  data.tar.gz: d34f68d51b328de8fb18cda83252d381c917de23b5737dc5e073b0ecc0c3939639c85b20b89dcabd0871343d5b3b7ea6918f2b268412213c951ee70c73a8321c

data/CHANGELOG.md

@@ -0,0 +1,73 @@
+## 4.2.6
+  - Removed JRuby check when using FIFOs [#75](https://github.com/logstash-plugins/logstash-output-file/pull/75)
+
+## 4.2.5
+  - Fix a bug introduced in v4.2.4 where events on low-volume pipelines could remain unflushed for long periods when `flush_interval` was non-zero [#70](https://github.com/logstash-plugins/logstash-output-file/pull/70)
+
+## 4.2.4
+  - Fix a bug where flush interval was being called for each event when enabled [#67](https://github.com/logstash-plugins/logstash-output-file/pull/67)
+
+## 4.2.3
+  - Docs: Set the default_codec doc attribute.
+
+## 4.2.2
+  - Add feature `write_behavior` to the documentation #58
+
+## 4.2.1
+  - Bugfix: Move require of flores into the spec file instead of main file.rb
+
+## 4.2.0
+  - New `write_behavior` feature. Value can be "append" (default) or
+    "overwrite". If "append", events will be appended to the end of the file.
+    If "overwrite", the file will only contain the last event written.
+
+## 4.1.2
+  - Update gemspec summary
+
+## 4.1.1
+  - Fix some documentation issues
+
+## 4.1.0
+  - Remove obsolete option `message_format`
+
+## 4.0.1
+  - Move one log message from info to debug to avoid noise
+
+## 4.0.0
+  - Make 'message_format' option obsolete
+  - Use new Logsash 2.4/5.0 APIs for working batchwise and with shared concurrency
+
+## 3.0.2
+  - Relax constraint on logstash-core-plugin-api to >= 1.60 <= 2.99
+
+## 3.0.1
+  - Republish all the gems under jruby.
+## 3.0.0
+  - Update the plugin to the version 2.0 of the plugin api, this change is required for Logstash 5.0 compatibility. See https://github.com/elastic/logstash/issues/5141
+# 2.2.5
+  - Depend on logstash-core-plugin-api instead of logstash-core, removing the need to mass update plugins on major releases of logstash
+# 2.2.4
+  - New dependency requirements for logstash-core for the 5.0 release
+## 2.2.3
+  - Rename Dir.exists? to Dir.exist? to fix deprecation warning
+  - Allow setting dir and file permissions
+
+## 2.2.1
+ - Fixed specs to not depend on pipeline ordering
+
+## 2.2.1
+ - Fixed Time specs
+
+## 2.2.0
+ - Add support for codec, using **json_lines** as default codec to keep default behavior.
+   Ref: https://github.com/logstash-plugins/logstash-output-file/pull/9
+
+## 2.1.0
+ - Add create_if_deleted option to create a destination file in case it
+   was deleted by another agent in the machine. In case of being false
+   the system will add the incomming messages to the failure file.
+
+## 2.0.0
+ - Plugins were updated to follow the new shutdown semantic, this mainly allows Logstash to instruct input plugins to terminate gracefully,
+   instead of using Thread.raise on the plugins' threads. Ref: https://github.com/elastic/logstash/pull/3895
+ - Dependency on logstash-core update to 2.0

data/CONTRIBUTORS

@@ -0,0 +1,20 @@
+The following is a list of people who have contributed ideas, code, bug
+reports, or in general have helped logstash along its way.
+
+Contributors:
+* Colin Surprenant (colinsurprenant)
+* Ivan Babrou (bobrik)
+* John E. Vincent (lusis)
+* Jordan Sissel (jordansissel)
+* Kayla Green (MixMuffins)
+* Kurt Hurtado (kurtado)
+* Matt Gray (mattgray)
+* Pete Fritchman (fetep)
+* Philippe Weber (wiibaa)
+* Pier-Hugues Pellerin (ph)
+* Richard Pijnenburg (electrical)
+
+Note: If you've sent us patches, bug reports, or otherwise contributed to
+Logstash, and you aren't on the list above and want to be, please let us know
+and we'll make sure you're here. Contributions from folks like you are what make
+open source awesome.

data/Gemfile

@@ -0,0 +1,11 @@
+source 'https://rubygems.org'
+
+gemspec
+
+logstash_path = ENV["LOGSTASH_PATH"] || "../../logstash"
+use_logstash_source = ENV["LOGSTASH_SOURCE"] && ENV["LOGSTASH_SOURCE"].to_s == "1"
+
+if Dir.exist?(logstash_path) && use_logstash_source
+  gem 'logstash-core', :path => "#{logstash_path}/logstash-core"
+  gem 'logstash-core-plugin-api', :path => "#{logstash_path}/logstash-core-plugin-api"
+end

data/LICENSE

@@ -0,0 +1,13 @@
+Copyright (c) 2012-2018 Elasticsearch <http://www.elastic.co>
+
+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/NOTICE.TXT

@@ -0,0 +1,5 @@
+Elasticsearch
+Copyright 2012-2015 Elasticsearch
+
+This product includes software developed by The Apache Software
+Foundation (http://www.apache.org/).

data/README.md

@@ -0,0 +1,98 @@
+# Logstash Plugin
+
+[![Travis Build Status](https://travis-ci.org/logstash-plugins/logstash-output-file.svg)](https://travis-ci.org/logstash-plugins/logstash-output-file)
+
+This is a plugin for [Logstash](https://github.com/elastic/logstash).
+
+It is fully free and fully open source. The license is Apache 2.0, meaning you are pretty much free to use it however you want in whatever way.
+
+## Documentation
+
+Logstash provides infrastructure to automatically generate documentation for this plugin. We use the asciidoc format to write documentation so any comments in the source code will be first converted into asciidoc and then into html. All plugin documentation are placed under one [central location](http://www.elastic.co/guide/en/logstash/current/).
+
+- For formatting code or config example, you can use the asciidoc `[source,ruby]` directive
+- For more asciidoc formatting tips, see the excellent reference here https://github.com/elastic/docs#asciidoc-guide
+
+## Need Help?
+
+Need help? Try #logstash on freenode IRC or the https://discuss.elastic.co/c/logstash discussion forum.
+
+## Developing
+
+### 1. Plugin Developement and Testing
+
+#### Code
+- To get started, you'll need JRuby with the Bundler gem installed.
+
+- Create a new plugin or clone and existing from the GitHub [logstash-plugins](https://github.com/logstash-plugins) organization. We also provide [example plugins](https://github.com/logstash-plugins?query=example).
+
+- Install dependencies
+```sh
+bundle install
+```
+
+#### Test
+
+- Update your dependencies
+
+```sh
+bundle install
+```
+
+- Run tests
+
+```sh
+bundle exec rspec
+```
+
+### 2. Running your unpublished Plugin in Logstash
+
+#### 2.1 Run in a local Logstash clone
+
+- Edit Logstash `Gemfile` and add the local plugin path, for example:
+```ruby
+gem "logstash-filter-awesome", :path => "/your/local/logstash-filter-awesome"
+```
+- Install plugin
+```sh
+# Logstash 2.3 and higher
+bin/logstash-plugin install --no-verify
+
+# Prior to Logstash 2.3
+bin/plugin install --no-verify
+
+```
+- Run Logstash with your plugin
+```sh
+bin/logstash -e 'filter {awesome {}}'
+```
+At this point any modifications to the plugin code will be applied to this local Logstash setup. After modifying the plugin, simply rerun Logstash.
+
+#### 2.2 Run in an installed Logstash
+
+You can use the same **2.1** method to run your plugin in an installed Logstash by editing its `Gemfile` and pointing the `:path` to your local plugin development directory or you can build the gem and install it using:
+
+- Build your plugin gem
+```sh
+gem build logstash-filter-awesome.gemspec
+```
+- Install the plugin from the Logstash home
+```sh
+# Logstash 2.3 and higher
+bin/logstash-plugin install --no-verify
+
+# Prior to Logstash 2.3
+bin/plugin install --no-verify
+
+```
+- Start Logstash and proceed to test the plugin
+
+## Contributing
+
+All contributions are welcome: ideas, patches, documentation, bug reports, complaints, and even something you drew up on a napkin.
+
+Programming is not a required skill. Whatever you've seen about open source and maintainers or community members  saying "send patches or die" - you will not see that here.
+
+It is more important to the community that you are able to contribute.
+
+For more information about contributing, see the [CONTRIBUTING](https://github.com/elastic/logstash/blob/master/CONTRIBUTING.md) file.

data/docs/index.asciidoc

@@ -0,0 +1,147 @@
+:plugin: file
+:type: output
+:default_codec: json_lines
+
+///////////////////////////////////////////
+START - GENERATED VARIABLES, DO NOT EDIT!
+///////////////////////////////////////////
+:version: %VERSION%
+:release_date: %RELEASE_DATE%
+:changelog_url: %CHANGELOG_URL%
+:include_path: ../../../../logstash/docs/include
+///////////////////////////////////////////
+END - GENERATED VARIABLES, DO NOT EDIT!
+///////////////////////////////////////////
+
+[id="plugins-{type}s-{plugin}"]
+
+=== File output plugin
+
+include::{include_path}/plugin_header.asciidoc[]
+
+==== Description
+
+This output writes events to files on disk. You can use fields
+from the event as parts of the filename and/or path.
+
+By default, this output writes one event per line in **json** format.
+You can customise the line format using the `line` codec like
+[source,ruby]
+output {
+ file {
+   path => ...
+   codec => line { format => "custom format: %{message}"}
+ }
+}
+
+[id="plugins-{type}s-{plugin}-options"]
+==== File Output Configuration Options
+
+This plugin supports the following configuration options plus the <<plugins-{type}s-{plugin}-common-options>> described later.
+
+[cols="<,<,<",options="header",]
+|=======================================================================
+|Setting |Input type|Required
+| <<plugins-{type}s-{plugin}-create_if_deleted>> |<<boolean,boolean>>|No
+| <<plugins-{type}s-{plugin}-dir_mode>> |<<number,number>>|No
+| <<plugins-{type}s-{plugin}-file_mode>> |<<number,number>>|No
+| <<plugins-{type}s-{plugin}-filename_failure>> |<<string,string>>|No
+| <<plugins-{type}s-{plugin}-flush_interval>> |<<number,number>>|No
+| <<plugins-{type}s-{plugin}-gzip>> |<<boolean,boolean>>|No
+| <<plugins-{type}s-{plugin}-path>> |<<string,string>>|Yes
+| <<plugins-{type}s-{plugin}-write_behavior>> |<<string,string>>|No
+|=======================================================================
+
+Also see <<plugins-{type}s-{plugin}-common-options>> for a list of options supported by all
+output plugins.
+
+&nbsp;
+
+[id="plugins-{type}s-{plugin}-create_if_deleted"]
+===== `create_if_deleted`
+
+  * Value type is <<boolean,boolean>>
+  * Default value is `true`
+
+If the configured file is deleted, but an event is handled by the plugin,
+the plugin will recreate the file. Default => true
+
+[id="plugins-{type}s-{plugin}-dir_mode"]
+===== `dir_mode`
+
+  * Value type is <<number,number>>
+  * Default value is `-1`
+
+Dir access mode to use. Note that due to the bug in jruby system umask
+is ignored on linux: https://github.com/jruby/jruby/issues/3426
+Setting it to -1 uses default OS value.
+Example: `"dir_mode" => 0750`
+
+[id="plugins-{type}s-{plugin}-file_mode"]
+===== `file_mode`
+
+  * Value type is <<number,number>>
+  * Default value is `-1`
+
+File access mode to use. Note that due to the bug in jruby system umask
+is ignored on linux: https://github.com/jruby/jruby/issues/3426
+Setting it to -1 uses default OS value.
+Example: `"file_mode" => 0640`
+
+[id="plugins-{type}s-{plugin}-filename_failure"]
+===== `filename_failure`
+
+  * Value type is <<string,string>>
+  * Default value is `"_filepath_failures"`
+
+If the generated path is invalid, the events will be saved
+into this file and inside the defined path.
+
+[id="plugins-{type}s-{plugin}-flush_interval"]
+===== `flush_interval`
+
+  * Value type is <<number,number>>
+  * Default value is `2`
+
+Flush interval (in seconds) for flushing writes to log files.
+0 will flush on every message.
+
+[id="plugins-{type}s-{plugin}-gzip"]
+===== `gzip`
+
+  * Value type is <<boolean,boolean>>
+  * Default value is `false`
+
+Gzip the output stream before writing to disk.
+
+[id="plugins-{type}s-{plugin}-path"]
+===== `path`
+
+  * This is a required setting.
+  * Value type is <<string,string>>
+  * There is no default value for this setting.
+
+The path to the file to write. Event fields can be used here,
+like `/var/log/logstash/%{host}/%{application}`
+One may also utilize the path option for date-based log
+rotation via the joda time format. This will use the event
+timestamp.
+E.g.: `path => "./test-%{+YYYY-MM-dd}.txt"` to create
+`./test-2013-05-29.txt`
+
+If you use an absolute path you cannot start with a dynamic string.
+E.g: `/%{myfield}/`, `/test-%{myfield}/` are not valid paths
+
+[id="plugins-{type}s-{plugin}-write_behavior"]
+===== `write_behavior`
+
+  * Value type is <<string,string>>
+  * Default value is `append`
+
+If `append`, the file will be opened for appending and each new event will be written at the end of the file.
+If `overwrite`, the file will be truncated before writing and only the most recent event will appear in the file.
+
+[id="plugins-{type}s-{plugin}-common-options"]
+include::{include_path}/{type}.asciidoc[]
+
+:default_codec!:

data/lib/logstash/outputs/googlecloudstorage.rb

@@ -0,0 +1,475 @@
+# encoding: utf-8
+require "logstash/namespace"
+require "logstash/outputs/base"
+require "logstash/json"
+require "logstash/errors"
+require "zlib"
+
+# This output writes events to files on disk. You can use fields
+# from the event as parts of the filename and/or path.
+#
+# By default, this output writes one event per line in **json** format.
+# You can customise the line format using the `line` codec like
+# [source,ruby]
+# output {
+#  file {
+#    path => ...
+#    codec => line { format => "custom format: %{message}"}
+#  }
+# }
+class LogStash::Outputs::GoogleCloudStorage < LogStash::Outputs::Base
+  concurrency :shared
+
+  FIELD_REF = /%\{[^}]+\}/
+
+  config_name "googlecloudstorage"
+
+  attr_reader :failure_path
+  
+  # GCS bucket name, without "gs://" or any other prefix.
+  config :bucket, :validate => :string, :required => true
+
+  # GCS path to private key file.
+  config :key_path, :validate => :string, :required => true
+
+  # GCS private key password.
+  config :key_password, :validate => :string, :default => "notasecret"
+
+  # GCS service account.
+  config :service_account, :validate => :string, :required => true
+
+  # Log file prefix. Log file will follow the format:
+  config :log_file_prefix, :validate => :string, :default => "logstash_gcs"
+
+  # The event format you want to store in files. Defaults to plain text.
+  config :output_format, :validate => [ "json", "plain" ], :default => "plain"
+
+  # Time pattern for log file, defaults to hourly files.
+  # Must Time.strftime patterns: www.ruby-doc.org/core-2.0/Time.html#method-i-strftime
+  config :date_pattern, :validate => :string, :default => "%Y-%m-%dT%H:00"
+ 
+  # Gzip output stream when writing events to log files, set
+  # `Content-Type` to `application/gzip` instead of `text/plain`, and
+  # use file suffix `.log.gz` instead of `.log`.
+  config :gzip, :validate => :boolean, :default => false
+
+  # Gzip output stream when writing events to log files and set
+  # `Content-Encoding` to `gzip`.
+  config :gzip_content_encoding, :validate => :boolean, :default => false
+
+  # Uploader interval when uploading new files to GCS. Adjust time based
+  # on your time pattern (for example, for hourly files, this interval can be
+  # around one hour).
+
+  # The path to the file to write. Event fields can be used here,
+  # like `/var/log/logstash/%{host}/%{application}`
+  # One may also utilize the path option for date-based log
+  # rotation via the joda time format. This will use the event
+  # timestamp.
+  # E.g.: `path => "./test-%{+YYYY-MM-dd}.txt"` to create
+  # `./test-2013-05-29.txt`
+  #
+  # If you use an absolute path you cannot start with a dynamic string.
+  # E.g: `/%{myfield}/`, `/test-%{myfield}/` are not valid paths
+  config :path, :validate => :string, :required => true
+
+  # Flush interval (in seconds) for flushing writes to log files.
+  # 0 will flush on every message.
+  config :flush_interval, :validate => :number, :default => 100
+
+  # Gzip the output stream before writing to disk.
+  config :gzip, :validate => :boolean, :default => false
+
+  # Should the hostname be included in the file name?
+  config :include_hostname, :validate => :boolean, :default => true
+
+  # If the generated path is invalid, the events will be saved
+  # into this file and inside the defined path.
+  config :filename_failure, :validate => :string, :default => '_filepath_failures'
+
+  # If the configured file is deleted, but an event is handled by the plugin,
+  # the plugin will recreate the file. Default => true
+  config :create_if_deleted, :validate => :boolean, :default => true
+
+  # Dir access mode to use. Note that due to the bug in jruby system umask
+  # is ignored on linux: https://github.com/jruby/jruby/issues/3426
+  # Setting it to -1 uses default OS value.
+  # Example: `"dir_mode" => 0750`
+  config :dir_mode, :validate => :number, :default => -1
+
+  # GoogleCloudStorage access mode to use. Note that due to the bug in jruby system umask
+  # is ignored on linux: https://github.com/jruby/jruby/issues/3426
+  # Setting it to -1 uses default OS value.
+  # Example: `"file_mode" => 0640`
+  config :file_mode, :validate => :number, :default => -1
+
+
+  # How should the file be written?
+  #
+  # If `append`, the file will be opened for appending and each new event will
+  # be written at the end of the file.
+  #
+  # If `overwrite`, the file will be truncated before writing and only the most
+  # recent event will appear in the file.
+  config :write_behavior, :validate => [ "overwrite", "append" ], :default => "append"
+
+  default :codec, "json_lines"
+
+  public
+  def register
+    require "fileutils" # For mkdir_p
+
+    @files = {}
+    @io_mutex = Mutex.new
+
+    @path = File.expand_path(path)
+
+    validate_path
+
+    if path_with_field_ref?
+      @file_root = extract_file_root
+    else
+      @file_root = File.dirname(path)
+    end
+    @failure_path = File.join(@file_root, @filename_failure)
+
+    @flush_interval = @flush_interval.to_i
+    if @flush_interval > 0
+      @flusher = Interval.start(@flush_interval, -> { flush_pending_files })
+    end
+
+    @content_type = @gzip ? 'application/gzip' : 'text/plain'
+    @content_encoding = @gzip_content_encoding ? 'gzip' : 'identity'
+
+    @last_stale_cleanup_cycle = Time.now
+    @stale_cleanup_interval = 10
+  end # def register
+
+  private
+  def validate_path
+    if (root_directory =~ FIELD_REF) != nil
+      @logger.error("GoogleCloudStorage: The starting part of the path should not be dynamic.", :path => @path)
+      raise LogStash::ConfigurationError.new("The starting part of the path should not be dynamic.")
+    end
+  end
+
+  def initialize_google_client
+    require "google/api_client"
+    require "openssl"
+
+    @client = Google::APIClient.new(:application_name => 'Logstash Google Cloud Storage output plugin', :application_version => '0.1')
+    @storage = @client.discovered_api('storage', 'v1')
+
+    key = Google::APIClient::PKCS12.load_key(@key_path, @key_password)
+    service_account = Google::APIClient::JWTAsserter.new(@service_account, 'https://www.googleapis.com/auth/devstorage.read_write', key)
+    @client.authorization = service_account.authorize
+  end
+
+  # Uploads a local file to the configured bucket.
+  def upload_object(filename)
+    begin
+      @logger.debug("GCS: upload object.", :filename => filename)
+
+      media = Google::APIClient::UploadIO.new(filename, @content_type)
+      metadata_insert_result = @client.execute(:api_method => @storage.objects.insert,
+                                               :parameters => {
+                                                 'uploadType' => 'multipart',
+                                                 'bucket' => @bucket,
+                                                 'contentEncoding' => @content_encoding,
+                                                 'name' => File.basename(filename)
+                                               },
+                                               :body_object => {contentType: @content_type},
+                                               :media => media)
+      contents = metadata_insert_result.data
+      @logger.debug("GCS: multipart insert",
+                    :object => contents.name,
+                    :self_link => contents.self_link)
+    rescue => e
+      @logger.error("GCS: failed to upload file", :exception => e)
+      # TODO(rdc): limit retries?
+      sleep 1
+      retry
+    end
+  end
+
+  private
+  def root_directory
+    parts = @path.split(File::SEPARATOR).select { |item| !item.empty?  }
+    if Gem.win_platform?
+      # First part is the drive letter
+      parts[1]
+    else
+      parts.first
+    end
+  end
+
+  public
+  def multi_receive_encoded(events_and_encoded)
+    encoded_by_path = Hash.new {|h,k| h[k] = []}
+
+    events_and_encoded.each do |event,encoded|
+      file_output_path = event_path(event)
+      encoded_by_path[file_output_path] << encoded
+    end
+
+    @io_mutex.synchronize do
+      encoded_by_path.each do |path,chunks|
+        fd = open(path)
+        if @write_behavior == "overwrite"
+          fd.truncate(0)
+          fd.seek(0, IO::SEEK_SET)
+          fd.write(chunks.last)
+        else
+          # append to the file
+          chunks.each {|chunk| fd.write(chunk + "\n") }
+        end
+        fd.flush unless @flusher && @flusher.alive?
+        upload_object(fd.path) unless @flusher && @flusher.alive?
+      end
+
+      close_stale_files
+    end
+  end # def receive
+
+  public
+  def close
+    @flusher.stop unless @flusher.nil?
+    @io_mutex.synchronize do
+      @logger.debug("Close: closing files")
+
+      @files.each do |path, fd|
+        begin
+          fd.close
+          @logger.debug("Closed file #{path}", :fd => fd)
+        rescue Exception => e
+          @logger.error("Exception while flushing and closing files.", :exception => e)
+        end
+      end
+    end
+  end
+
+  private
+  def inside_file_root?(log_path)
+    target_file = File.expand_path(log_path)
+    return target_file.start_with?("#{@file_root.to_s}/")
+  end
+
+  private
+  def event_path(event)
+    file_output_path = generate_filepath(event)
+    if path_with_field_ref? && !inside_file_root?(file_output_path)
+      @logger.warn("GoogleCloudStorage: the event tried to write outside the files root, writing the event to the failure file",  :event => event, :filename => @failure_path)
+      file_output_path = @failure_path
+    elsif !@create_if_deleted && deleted?(file_output_path)
+      file_output_path = @failure_path
+    end
+    @logger.debug("GoogleCloudStorage, writing event to file.", :filename => file_output_path)
+
+    file_output_path
+  end
+
+  private
+  def generate_filepath(event)
+    event.sprintf(@path)
+  end
+
+  private
+  def path_with_field_ref?
+    path =~ FIELD_REF
+  end
+
+  private
+  def extract_file_root
+    parts = File.expand_path(path).split(File::SEPARATOR)
+    parts.take_while { |part| part !~ FIELD_REF }.join(File::SEPARATOR)
+  end
+
+  # the back-bone of @flusher, our periodic-flushing interval.
+  private
+  def flush_pending_files
+    @io_mutex.synchronize do
+      @logger.debug("Starting flush cycle")
+
+      @files.each do |path, fd|
+        @logger.debug("Flushing file", :path => path, :fd => fd)
+        fd.flush
+        upload_object(fd.path)
+      end
+    end
+  rescue => e
+    # squash exceptions caught while flushing after logging them
+    @logger.error("Exception flushing files", :exception => e.message, :backtrace => e.backtrace)
+  end
+
+  # every 10 seconds or so (triggered by events, but if there are no events there's no point closing files anyway)
+  private
+  def close_stale_files
+    now = Time.now
+    return unless now - @last_stale_cleanup_cycle >= @stale_cleanup_interval
+
+    @logger.debug("Starting stale files cleanup cycle", :files => @files)
+    inactive_files = @files.select { |path, fd| not fd.active }
+    @logger.debug("%d stale files found" % inactive_files.count, :inactive_files => inactive_files)
+    inactive_files.each do |path, fd|
+      @logger.info("Closing file %s" % path)
+      fd.close
+      @files.delete(path)
+    end
+    # mark all files as inactive, a call to write will mark them as active again
+    @files.each { |path, fd| fd.active = false }
+    @last_stale_cleanup_cycle = now
+  end
+
+  private
+  def cached?(path)
+    @files.include?(path) && !@files[path].nil?
+  end
+
+  private
+  def deleted?(path)
+    !File.exist?(path)
+  end
+
+  private
+  def open(path)
+    if !deleted?(path) && cached?(path)
+      return @files[path]
+    end
+
+    if deleted?(path)
+      if @create_if_deleted
+        @logger.debug("Required path was deleted, creating the file again", :path => path)
+        @files.delete(path)
+      else
+        return @files[path] if cached?(path)
+      end
+    end
+
+    @logger.info("Opening file", :path => path)
+
+    dir = File.dirname(path)
+    if !Dir.exist?(dir)
+      @logger.info("Creating directory", :directory => dir)
+      if @dir_mode != -1
+        FileUtils.mkdir_p(dir, :mode => @dir_mode)
+      else
+        FileUtils.mkdir_p(dir)
+      end
+    end
+
+    # work around a bug opening fifos (bug JRUBY-6280)
+    stat = File.stat(path) rescue nil
+    if stat && stat.ftype == "fifo"
+      fd = java.io.FileWriter.new(java.io.File.new(path))
+    else
+      if @file_mode != -1
+        fd = File.new(path, "a+", @file_mode)
+      else
+        fd = File.new(path, "a+")
+      end
+    end
+    if gzip
+      fd = Zlib::GzipWriter.new(fd)
+    end
+    @files[path] = IOWriter.new(fd)
+  end
+
+  ##
+  # Bare-bones utility for running a block of code at an interval.
+  #
+  class Interval
+    ##
+    # Initializes a new Interval with the given arguments and starts it before returning it.
+    #
+    # @param interval [Integer] (see: Interval#initialize)
+    # @param procsy [#call] (see: Interval#initialize)
+    #
+    # @return [Interval]
+    #
+    def self.start(interval, procsy)
+      self.new(interval, procsy).tap(&:start)
+    end
+
+    ##
+    # @param interval [Integer]: time in seconds to wait between calling the given proc
+    # @param procsy [#call]: proc or lambda to call periodically; must not raise exceptions.
+    def initialize(interval, procsy)
+      @interval = interval
+      @procsy = procsy
+
+      require 'thread' # Mutex, ConditionVariable, etc.
+      @mutex = Mutex.new
+      @sleeper = ConditionVariable.new
+    end
+
+    ##
+    # Starts the interval, or returns if it has already been started.
+    #
+    # @return [void]
+    def start
+      @mutex.synchronize do
+        return if @thread && @thread.alive?
+
+        @thread = Thread.new { run }
+      end
+    end
+
+    ##
+    # Stop the interval.
+    # Does not interrupt if execution is in-progress.
+    def stop
+      @mutex.synchronize do
+        @stopped = true
+      end
+
+      @thread && @thread.join
+    end
+
+    ##
+    # @return [Boolean]
+    def alive?
+      @thread && @thread.alive?
+    end
+
+    private
+
+    def run
+      @mutex.synchronize do
+        loop do
+          @sleeper.wait(@mutex, @interval)
+          break if @stopped
+
+          @procsy.call
+        end
+      end
+    ensure
+      @sleeper.broadcast
+    end
+  end # class LogStash::Outputs::File::Interval
+end # class LogStash::Outputs::File
+
+# wrapper class
+class IOWriter
+  def initialize(io)
+    @io = io
+  end
+  def write(*args)
+    @io.write(*args)
+    @active = true
+  end
+  def flush
+    @io.flush
+    if @io.class == Zlib::GzipWriter
+      @io.to_io.flush
+    end
+  end
+  def method_missing(method_name, *args, &block)
+    if @io.respond_to?(method_name)
+
+      @io.send(method_name, *args, &block)
+    else
+      super
+    end
+  end
+  attr_accessor :active
+end