Merge pull request #742 from estolfo/benchmarking-1

Driver Benchmark tests

Author: estolfo

Gemfile

@@ -10,6 +10,8 @@ group :development, :testing do
   gem 'rspec', '~> 3.0'
   gem 'mime-types', '~> 1.25'
   gem 'httparty'
+  gem 'yajl-ruby', require: 'yajl', platforms: :mri
+  gem 'celluloid', platforms: :mri
 end
 
 group :development do

Rakefile

@@ -41,3 +41,158 @@ namespace :docs do
     system "yardoc -o #{out} --title mongo-#{Mongo::VERSION}"
   end
 end
+
+require_relative "profile/benchmarking"
+
+# Some require data files, available from the drivers team. See the comments above each task for details."
+namespace :benchmark do
+  desc "Run the driver benchmark tests."
+
+  namespace :micro do
+    desc "Run the common driver micro benchmarking tests"
+
+    namespace :flat do
+      desc "Benchmarking for flat bson documents."
+
+      # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called flat_bson.json.
+      task :encode do
+        puts "MICRO BENCHMARK:: FLAT:: ENCODE"
+        Mongo::Benchmarking::Micro.run(:flat, :encode)
+      end
+
+      # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called flat_bson.json.
+      task :decode do
+        puts "MICRO BENCHMARK:: FLAT:: DECODE"
+        Mongo::Benchmarking::Micro.run(:flat, :decode)
+      end
+    end
+
+    namespace :deep do
+      desc "Benchmarking for deep bson documents."
+
+      # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called deep_bson.json.
+      task :encode do
+        puts "MICRO BENCHMARK:: DEEP:: ENCODE"
+        Mongo::Benchmarking::Micro.run(:deep, :encode)
+      end
+
+      # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called deep_bson.json.
+      task :decode do
+        puts "MICRO BENCHMARK:: DEEP:: DECODE"
+        Mongo::Benchmarking::Micro.run(:deep, :decode)
+      end
+    end
+
+    namespace :full do
+      desc "Benchmarking for full bson documents."
+
+      # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called full_bson.json.
+      task :encode do
+        puts "MICRO BENCHMARK:: FULL:: ENCODE"
+        Mongo::Benchmarking::Micro.run(:full, :encode)
+      end
+
+      # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called full_bson.json.
+      task :decode do
+        puts "MICRO BENCHMARK:: FULL:: DECODE"
+        Mongo::Benchmarking::Micro.run(:full, :decode)
+      end
+    end
+  end
+
+  namespace :single_doc do
+    desc "Run the common driver single-document benchmarking tests"
+    task :command do
+      puts "SINGLE DOC BENCHMARK:: COMMAND"
+      Mongo::Benchmarking::SingleDoc.run(:command)
+    end
+
+    # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called TWEET.json.
+    task :find_one do
+      puts "SINGLE DOC BENCHMARK:: FIND ONE BY ID"
+      Mongo::Benchmarking::SingleDoc.run(:find_one)
+    end
+
+    # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called SMALL_DOC.json.
+    task :insert_one_small do
+      puts "SINGLE DOC BENCHMARK:: INSERT ONE SMALL DOCUMENT"
+      Mongo::Benchmarking::SingleDoc.run(:insert_one_small)
+    end
+
+    # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called LARGE_DOC.json.
+    task :insert_one_large do
+      puts "SINGLE DOC BENCHMARK:: INSERT ONE LARGE DOCUMENT"
+      Mongo::Benchmarking::SingleDoc.run(:insert_one_large)
+    end
+  end
+
+  namespace :multi_doc do
+    desc "Run the common driver multi-document benchmarking tests"
+
+    # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called TWEET.json.
+    task :find_many do
+      puts "MULTI DOCUMENT BENCHMARK:: FIND MANY"
+      Mongo::Benchmarking::MultiDoc.run(:find_many)
+    end
+
+    # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called SMALL_DOC.json.
+    task :bulk_insert_small do
+      puts "MULTI DOCUMENT BENCHMARK:: BULK INSERT SMALL"
+      Mongo::Benchmarking::MultiDoc.run(:bulk_insert_small)
+    end
+
+    # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called LARGE_DOC.json.
+    task :bulk_insert_large do
+      puts "MULTI DOCUMENT BENCHMARK:: BULK INSERT LARGE"
+      Mongo::Benchmarking::MultiDoc.run(:bulk_insert_large)
+    end
+
+    # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called GRIDFS_LARGE.
+    task :gridfs_upload do
+      puts "MULTI DOCUMENT BENCHMARK:: GRIDFS UPLOAD"
+      Mongo::Benchmarking::MultiDoc.run(:gridfs_upload)
+    end
+
+    # Requirement: A file in Mongo::Benchmarking::DATA_PATH, called GRIDFS_LARGE.
+    task :gridfs_download do
+      puts "MULTI DOCUMENT BENCHMARK:: GRIDFS DOWNLOAD"
+      Mongo::Benchmarking::MultiDoc.run(:gridfs_download)
+    end
+  end
+
+  namespace :parallel do
+    desc "Run the common driver paralell ETL benchmarking tests"
+
+    # Requirement: A directory in Mongo::Benchmarking::DATA_PATH, called LDJSON_MULTI,
+    # with the files used in this task.
+    task :import do
+      puts "PARALLEL ETL BENCHMARK:: IMPORT"
+      Mongo::Benchmarking::Parallel.run(:import)
+    end
+
+    # Requirement: A directory in Mongo::Benchmarking::DATA_PATH, called LDJSON_MULTI,
+    # with the files used in this task.
+    # Requirement: Another directory in "#{Mongo::Benchmarking::DATA_PATH}/LDJSON_MULTI"
+    # called 'output'.
+    task :export do
+      puts "PARALLEL ETL BENCHMARK:: EXPORT"
+      Mongo::Benchmarking::Parallel.run(:export)
+    end
+
+    # Requirement: A directory in Mongo::Benchmarking::DATA_PATH, called GRIDFS_MULTI,
+    # with the files used in this task.
+    task :gridfs_upload do
+      puts "PARALLEL ETL BENCHMARK:: GRIDFS UPLOAD"
+      Mongo::Benchmarking::Parallel.run(:gridfs_upload)
+    end
+
+    # Requirement: A directory in Mongo::Benchmarking::DATA_PATH, called GRIDFS_MULTI,
+    # with the files used in this task.
+    # Requirement: Another directory in "#{Mongo::Benchmarking::DATA_PATH}/GRIDFS_MULTI"
+    # called 'output'.
+    task :gridfs_download do
+      puts "PARALLEL ETL BENCHMARK:: GRIDFS DOWNLOAD"
+      Mongo::Benchmarking::Parallel.run(:gridfs_download)
+    end
+  end
+end

profile/benchmarking.rb

@@ -0,0 +1,128 @@
+# Copyright (C) 2015 MongoDB, Inc.
+#
+# 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.
+
+require 'benchmark'
+require_relative 'benchmarking/helper'
+require_relative 'benchmarking/micro'
+require_relative 'benchmarking/single_doc'
+require_relative 'benchmarking/multi_doc'
+require_relative 'benchmarking/parallel'
+
+module Mongo
+
+  # Module with all functionality for running driver benchmark tests.
+  #
+  # @since 2.2.3
+  module Benchmarking
+
+    extend self
+
+    # The current path.
+    #
+    # @return [ String ] The current path.
+    #
+    # @since 2.2.3
+    CURRENT_PATH = File.expand_path(File.dirname(__FILE__)).freeze
+
+    # The path to data files used in Benchmarking tests.
+    #
+    # @return [ String ] Path to Benchmarking test files.
+    #
+    # @since 2.2.3
+    DATA_PATH = [CURRENT_PATH, 'benchmarking', 'data'].join('/').freeze
+
+    # The file containing the single tweet document.
+    #
+    # @return [ String ] The file containing the tweet document.
+    #
+    # @since 2.2.3
+    TWEET_DOCUMENT_FILE = [DATA_PATH, 'TWEET.json'].join('/').freeze
+
+    # The file containing the single small document.
+    #
+    # @return [ String ] The file containing the small document.
+    #
+    # @since 2.2.3
+    SMALL_DOCUMENT_FILE = [DATA_PATH, 'SMALL_DOC.json'].join('/').freeze
+
+    # The file containing the single large document.
+    #
+    # @return [ String ] The file containing the large document.
+    #
+    # @since 2.2.3
+    LARGE_DOCUMENT_FILE = [DATA_PATH, 'LARGE_DOC.json'].join('/').freeze
+
+    # The file to upload when testing GridFS.
+    #
+    # @return [ String ] The file containing the GridFS test data.
+    #
+    # @since 2.2.3
+    GRIDFS_FILE = [DATA_PATH, 'GRIDFS_LARGE'].join('/').freeze
+
+    # The file path and base name for the LDJSON files.
+    #
+    # @return [ String ] The file path and base name for the LDJSON files.
+    #
+    # @since 2.2.3
+    LDJSON_FILE_BASE = [DATA_PATH, 'LDJSON_MULTI', 'LDJSON'].join('/').freeze
+
+    # The file path and base name for the outputted LDJSON files.
+    #
+    # @return [ String ] The file path and base name for the outputted LDJSON files.
+    #
+    # @since 2.2.3
+    LDJSON_FILE_OUTPUT_BASE = [DATA_PATH, 'LDJSON_MULTI', 'output', 'LDJSON'].join('/').freeze
+
+    # The file path and base name for the GRIDFS files to upload.
+    #
+    # @return [ String ] The file path and base name for the GRIDFS files to upload.
+    #
+    # @since 2.2.3
+    GRIDFS_MULTI_BASE = [DATA_PATH, 'GRIDFS_MULTI', 'file'].join('/').freeze
+
+    # The file path and base name for the outputted GRIDFS downloaded files.
+    #
+    # @return [ String ] The file path and base name for the outputted GRIDFS downloaded files.
+    #
+    # @since 2.2.3
+    GRIDFS_MULTI_OUTPUT_BASE = [DATA_PATH, 'GRIDFS_MULTI', 'output', 'file-output'].join('/').freeze
+
+    # The default number of test repetitions.
+    #
+    # @return [ Integer ] The number of test repetitions.
+    #
+    # @since 2.2.3
+    TEST_REPETITIONS = 100.freeze
+
+    # The number of default warmup repetitions of the test to do before
+    # recording times.
+    #
+    # @return [ Integer ] The default number of warmup repetitions.
+    #
+    # @since 2.2.3
+    WARMUP_REPETITIONS = 10.freeze
+
+    def tweet_document
+      Benchmarking.load_file(TWEET_DOCUMENT_FILE).first
+    end
+
+    def small_document
+      Benchmarking.load_file(SMALL_DOCUMENT_FILE).first
+    end
+
+    def large_document
+      Benchmarking.load_file(LARGE_DOCUMENT_FILE).first
+    end
+  end
+end

profile/benchmarking/helper.rb

@@ -0,0 +1,59 @@
+module Mongo
+
+  # Helper functions used by benchmarking tasks
+  module Benchmarking
+
+    extend self
+
+    # Load a json file and represent each document as a Hash.
+    #
+    # @example Load a file.
+    #   Benchmarking.load_file(file_name)
+    #
+    # @param [ String ] The file name.
+    #
+    # @return [ Array ] A list of extended-json documents.
+    #
+    # @since 2.2.3
+    def load_file(file_name)
+      File.open(file_name, "r") do |f|
+        f.each_line.collect do |line|
+          parse_json(line)
+        end
+      end
+    end
+
+    # Load a json document as a Hash and convert BSON-specific types.
+    # Replace the _id field as an BSON::ObjectId if it's represented as '$oid'.
+    #
+    # @example Parse a json document.
+    #   Benchmarking.parse_json(document)
+    #
+    # @param [ Hash ] The json document.
+    #
+    # @return [ Hash ] An extended-json document.
+    #
+    # @since 2.2.3
+    def parse_json(document)
+      JSON.parse(document).tap do |doc|
+        if doc['_id'] && doc['_id']['$oid']
+          doc['_id'] = BSON::ObjectId.from_string(doc['_id']['$oid'])
+        end
+      end
+    end
+
+    # Get the median of values in a list.
+    #
+    # @example Get the median.
+    #   Benchmarking.median(values)
+    #
+    # @param [ Array ] The values to get the median of.
+    #
+    # @return [ Numeric ] The median of the list.
+    #
+    # @since 2.2.3
+    def median(values)
+      values.sort![values.size/2-1]
+    end
+  end
+end