Simple benchmarks for query plan reuse.

Author: metaskills

Gemfile

@@ -19,6 +19,7 @@ group :development do
   gem 'rake', '>= 0.8.7'
   gem 'mocha', '0.9.8'
   gem 'shoulda', '2.10.3'
+  gem 'bench_press'
   platforms :mri_18 do
     gem 'ruby-prof', '0.9.1'
     gem 'ruby-debug', '0.10.3'

test/profile/query_plan_complex.rb

@@ -0,0 +1,68 @@
+=begin
+
+Query Plan Complex
+==================
+Author: Ken Collins  
+Date: May 22, 2011  
+Summary: Benchmark complex cached query plan reuse in SQL Server.  
+
+System Information
+------------------
+    Operating System:    Mac OS X 10.6.7 (10J869)
+    CPU:                 Quad-Core Intel Xeon 2.66 GHz
+    Processor Count:     4
+    Memory:              24 GB
+    ruby 1.8.7 (2011-02-18 patchlevel 334) [i686-darwin10.6.0], MBARI 0x6770, Ruby Enterprise Edition 2011.03
+
+"Simple - Query Plan Reuse" is up to 88% faster over  repetitions
+-----------------------------------------------------------------
+
+    Simple - Query Plan Reuse    0.230067014694214 secs    Fastest
+    Simple - Dynamic SQL         1.99195981025696  secs    88% Slower
+
+=end
+
+require 'rubygems'
+require 'bundler'
+Bundler.setup
+require 'tiny_tds'
+require 'bench_press'
+
+extend BenchPress
+
+author 'Ken Collins'
+summary 'Benchmark complex cached query plan reuse in SQL Server.'
+reps 500
+
+@client = TinyTds::Client.new :host => 'mc2008', :username => 'rails'
+
+
+measure "Simple - Dynamic SQL" do
+  sql = "
+    SELECT TOP (1) [companies].id 
+    FROM [companies] 
+    LEFT OUTER JOIN [companies] [clients_using_primary_keys_companies] ON [clients_using_primary_keys_companies].[firm_name] = [companies].[name] 
+    AND [clients_using_primary_keys_companies].[type] IN (N'Client', N'SpecialClient', N'VerySpecialClient') 
+    WHERE [companies].[type] IN (N'Firm') 
+    AND [companies].[id] = #{rand(1000000)} 
+    GROUP BY [companies].id 
+    ORDER BY MIN(clients_using_primary_keys_companies.name)"
+  @client.execute(sql).do
+end
+
+measure "Simple - Query Plan Reuse" do
+  sql = "
+    EXEC sp_executesql N'
+      SELECT TOP (1) [companies].id 
+      FROM [companies] 
+      LEFT OUTER JOIN [companies] [clients_using_primary_keys_companies] ON [clients_using_primary_keys_companies].[firm_name] = [companies].[name] 
+      AND [clients_using_primary_keys_companies].[type] IN (N''Client'', N''SpecialClient'', N''VerySpecialClient'') 
+      WHERE [companies].[type] IN (N''Firm'') 
+      AND [companies].[id] = @0 
+      GROUP BY [companies].id 
+      ORDER BY MIN(clients_using_primary_keys_companies.name)', 
+    N'@0 int', 
+    @0 = #{rand(1000000)}"
+  @client.execute(sql).do
+end
+

test/profile/query_plan_simple.rb

@@ -0,0 +1,47 @@
+=begin
+
+Query Plan Simple
+=================
+Author: Ken Collins  
+Date: May 22, 2011  
+Summary: Benchmark simple cached query plan reuse in SQL Server.  
+
+System Information
+------------------
+    Operating System:    Mac OS X 10.6.7 (10J869)
+    CPU:                 Quad-Core Intel Xeon 2.66 GHz
+    Processor Count:     4
+    Memory:              24 GB
+    ruby 1.8.7 (2011-02-18 patchlevel 334) [i686-darwin10.6.0], MBARI 0x6770, Ruby Enterprise Edition 2011.03
+
+"Simple - Query Plan Reuse" is up to 58% faster over  repetitions
+-----------------------------------------------------------------
+
+    Simple - Query Plan Reuse    0.20799994468689 secs    Fastest
+    Simple - Dynamic SQL         0.49638819694519 secs    58% Slower
+
+=end
+
+require 'rubygems'
+require 'bundler'
+Bundler.setup
+require 'tiny_tds'
+require 'bench_press'
+
+extend BenchPress
+
+author 'Ken Collins'
+summary 'Benchmark simple cached query plan reuse in SQL Server.'
+reps 500
+
+@client = TinyTds::Client.new :host => 'mc2008', :username => 'rails'
+
+
+measure "Simple - Dynamic SQL" do
+  @client.execute("SELECT TOP(1) * FROM [posts] WHERE [id] = #{rand(1000000)}").do
+end
+
+measure "Simple - Query Plan Reuse" do
+  @client.execute("EXEC sp_executesql N'SELECT TOP(1) * FROM [posts] WHERE [id] = @0', N'@0 int', @0 = #{rand(1000000)}").do
+end
+