Rake tasks to run tests or specs in parallel, to use multiple CPUs and speedup test runtime. more documentation and great illustrations

Setup

script/plugin install git://github.com/grosser/parallel_specs.git

Add <%= ENV['TEST_ENV_NUMBER'] %> to the database name for the test environment in config/database.yml,
it is '' for process 1, and '2' for process 2.

test:
  adapter: mysql
  database: xxx_test<%= ENV['TEST_ENV_NUMBER'] %>
  username: root

Create the databases mysql -u root create database xxx_test; #should normally exist... create database xxx_test2; ...

Run like hell :D rake parallel:prepare[2] #db:reset for 2 databases

rake parallel:spec[1] --> 1 cpu  --> 86 seconds
#OR for Test::Unit
rake parallel:test[1]

rake parallel:spec    --> 2 cpus --> 47 seconds
rake parallel:spec[4] --> 4 cpus --> 26 seconds
...

Just some subfolders please (e.g. set up one integration server to check each subfolder) rake parallel:spec[2,models] rake parallel:test[2,something/else]

partial paths are OK too...
rake parallel:test[2,functional] == rake parallel:test[2,fun]

Example output

2 processes for 210 specs, ~ 105 specs per process
... test output ...

Results:
877 examples, 0 failures, 11 pending
843 examples, 0 failures, 1 pending

Took 29.925333 seconds

TIPS

• 'script/spec_server' or spork do not work in parallel
• ./script/generate rspec if you are running rspec from gems (this plugin uses script/spec which may fail if rspec files are outdated)
• with zsh this would be rake "parallel:prepare[3]"

TODO

• build parallel:bootstrap idea/basics
• make jRuby compatible basics
• make windows compatible (does anyone care ?)

Authors

inspired by pivotal labs

###Contributors

• Joakim Kolsjö -- joakim.kolsjo<$at$>gmail.com
• Jason Morrison -- jason.p.morrison<$at$>gmail.com
• Tchandy -- tchandy<$at$>gmail.com

Michael Grosser
grosser.michael@gmail.com
Hereby placed under public domain, do what you want, just do not hold me accountable...