From 4cc24461b32500a0dc4abec6172ab9fb91fd0406 Mon Sep 17 00:00:00 2001
From: Takashi Kokubun <takashikkbn@gmail.com>
Date: Thu, 27 Jul 2023 10:14:10 -0700
Subject: [PATCH] Add fluentd benchmark (#243)

* Add fluentd benchmark

* Add more variants to error messages
---
 benchmarks.yml                  |  2 ++
 benchmarks/fluentd/Gemfile      |  5 ++++
 benchmarks/fluentd/Gemfile.lock | 46 +++++++++++++++++++++++++++++++++
 benchmarks/fluentd/benchmark.rb | 30 +++++++++++++++++++++
 4 files changed, 83 insertions(+)
 create mode 100644 benchmarks/fluentd/Gemfile
 create mode 100644 benchmarks/fluentd/Gemfile.lock
 create mode 100644 benchmarks/fluentd/benchmark.rb

diff --git a/benchmarks.yml b/benchmarks.yml
index 65237e79..0a2a7bae 100644
--- a/benchmarks.yml
+++ b/benchmarks.yml
@@ -49,6 +49,8 @@ etanni:
   desc: etanni is an older, extremely simple template-lang format that basically turns your template into an "eval" with a lot of heredocs.
 fannkuchredux:
   desc: fannkuchredux from the Computer Language Benchmarks Game.
+fluentd:
+  desc: fluentd is a log collector, which parses logs in a server and forwards them to various destinations.
 ruby-json:
   desc: an optimized version of the json_pure gem's pure Ruby JSON parser.
 lee:
diff --git a/benchmarks/fluentd/Gemfile b/benchmarks/fluentd/Gemfile
new file mode 100644
index 00000000..3f7549a1
--- /dev/null
+++ b/benchmarks/fluentd/Gemfile
@@ -0,0 +1,5 @@
+source 'https://rubygems.org'
+git_source(:github) { |repo| "https://github.com/#{repo}.git" }
+
+gem 'fluentd'
+gem 'cool.io', github: 'k0kubun/cool.io', branch: 'ruby-3-3'
diff --git a/benchmarks/fluentd/Gemfile.lock b/benchmarks/fluentd/Gemfile.lock
new file mode 100644
index 00000000..1dda7d59
--- /dev/null
+++ b/benchmarks/fluentd/Gemfile.lock
@@ -0,0 +1,46 @@
+GIT
+  remote: https://github.com/k0kubun/cool.io.git
+  revision: 61e3caf8b961fb1a4d46b367eb58ebcddb45da4a
+  branch: ruby-3-3
+  specs:
+    cool.io (1.7.1)
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    concurrent-ruby (1.2.2)
+    fluentd (1.16.2)
+      bundler
+      cool.io (>= 1.4.5, < 2.0.0)
+      http_parser.rb (>= 0.5.1, < 0.9.0)
+      msgpack (>= 1.3.1, < 2.0.0)
+      serverengine (>= 2.3.2, < 3.0.0)
+      sigdump (~> 0.2.5)
+      strptime (>= 0.2.4, < 1.0.0)
+      tzinfo (>= 1.0, < 3.0)
+      tzinfo-data (~> 1.0)
+      webrick (~> 1.4)
+      yajl-ruby (~> 1.0)
+    http_parser.rb (0.8.0)
+    msgpack (1.7.2)
+    serverengine (2.3.2)
+      sigdump (~> 0.2.2)
+    sigdump (0.2.5)
+    strptime (0.2.5)
+    tzinfo (2.0.6)
+      concurrent-ruby (~> 1.0)
+    tzinfo-data (1.2023.3)
+      tzinfo (>= 1.0.0)
+    webrick (1.8.1)
+    yajl-ruby (1.4.3)
+
+PLATFORMS
+  arm64-darwin-22
+  x86_64-linux
+
+DEPENDENCIES
+  cool.io!
+  fluentd
+
+BUNDLED WITH
+   2.4.1
diff --git a/benchmarks/fluentd/benchmark.rb b/benchmarks/fluentd/benchmark.rb
new file mode 100644
index 00000000..aee1908d
--- /dev/null
+++ b/benchmarks/fluentd/benchmark.rb
@@ -0,0 +1,30 @@
+# frozen_string_literal: true
+
+require 'harness'
+
+Dir.chdir(__dir__)
+use_gemfile
+
+require 'fluent/engine'
+require 'fluent/parser'
+
+# Prepare a fixture
+time = Time.new(2023, 7, 27, 9, 00, 00)
+errors = [
+  "Skipping user-data validation. No user-data found.",
+  "Python version change detected. Purging cache",
+  "No instance datasource found.",
+  "No kernel command line url found.",
+  "No local datasource found",
+]
+ltsv = 1000.times.map { |i| "time:#{time + i}	module:main.py	level:DEBUG	message:#{errors.sample}\n" }.join
+ltsv *= 1000
+
+# Prepare an LTSV parser
+parser = Fluent::Plugin::LabeledTSVParser.new
+parser.configure(Fluent::Config::Element.new('parse', '', {}, []))
+
+# Benchmark the `<parse>@type ltsv</parse>` use case
+run_benchmark(10) do
+  parser.parse(ltsv) {}
+end