Skip to content

kigster/performance-compare

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Performance::Compare

This gem can be used to compare various implementations of the same algorithm using a single-threaded Ruby process, multi-threaded process, or multi-process pool of workers.

Concurrency here is achieved via the gem parallel which supports both thread and process pools.

Provided Example

In the provided example (see the lib/algorithms folder) we provide two implementations of the sha1sum algorithm:

  1. Using Ruby's Digest::SHA1.digest implementation
  2. By shelling out to use the sha1sum binary supplied with GNU Core Utils.

Installation

git clone https://github.com/kigster/performance-compare
cd performance-compare
gem install bundler
bundle

Usage

You run this tool via exe/compare N1, N2, ... command line syntax. Each number corresponds to the total number of iterations for each algorithm. This is done so that algorithms that are vastly different in speed can still be effectively compared.

❯ exe/compare 10000 1000
┌──────────────────────────────────────────────────────────────────────────────────────────────────┐
│                                                                                                  │
│ Number of CPU Cores:                 16                                                          │
│ Method under the test:               sha1sum                                                     │
│ Number of iterations per algorithm:  [10000, 1000]                                               │
│ List of algorithms to compare:       ["Algorithms::Sha1Sum::Ruby", "Algorithms::Sha1Sum::Fork"]  │
│                                                                                                  │
└──────────────────────────────────────────────────────────────────────────────────────────────────┘
                             user     system      total        real
┌──────────────────────────────────────────────────────────────────────────────────────────────────┐
│Testing Implementation: Algorithms::Sha1Sum::Ruby, 10000 iterations                               │
└──────────────────────────────────────────────────────────────────────────────────────────────────┘
sha1sum |    threads |  1  0.062405   0.001820   0.064225 (  0.065957)
sha1sum |    threads | 16  0.067567   0.003203   0.070770 (  0.068327)
sha1sum |  processes |  1  0.077446   0.039013   0.282787 (  0.251656)
sha1sum |  processes | 16  0.140628   0.157827   0.765949 (  0.216212)
┌──────────────────────────────────────────────────────────────────────────────────────────────────┐
│Testing Implementation: Algorithms::Sha1Sum::Fork, 1000 iterations                                │
└──────────────────────────────────────────────────────────────────────────────────────────────────┘
sha1sum |    threads |  1  0.292745   1.256383  12.996616 ( 16.418338)
sha1sum |    threads | 16  0.351036   1.789147  22.926779 (  3.586996)
sha1sum |  processes |  1  0.031150   0.014962  12.794822 ( 15.831681)
sha1sum |  processes | 16  0.036771   0.030601  18.085052 (  7.244181)

Intepreting Results

The above results show

Development

After checking out the repo, run bin/setup to install dependencies. Then, run rake spec to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

Contributing

Bug reports and pull requests are welcome on GitHub at https://github.com/kigster/performance-compare.

About

A simple gem that can be used to compare algorithm implementations in Ruby using a single thread, a thread pool, or a process pool.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors