grunt-parallel-spec-runner

Plugin used to configure and launch multiple spec runners in parallel.

Getting Started

Install the module with: npm install grunt-parallel-spec-runner

The "grunt-parallel-spec-runner" task

Overview

In your project's Gruntfile, add a section named grunt-parallel-spec-runner to the data object passed into grunt.initConfig().

grunt.initConfig
  parallel_spec_runner:
    your_target:
        options:
           specs:
           excludedSpecs:
           isolatedSpecs:

Options

options.specs

Type: String|Array

Your Jasmine specs.

options.excludedSpecs

Type: String|Array

Specifically excludes any spec files that were included via options.spec.

options.isolatedSpecs

Type: String|Array

Specifically isolate any spec files that were included via options.spec. This is useful to isolate specs that do not play well with others, in addition it can be faster to isolate very large spec files so that they run in their own webdriver/browser so they don't block other specs that run quickly. Spec files that take more than a couple of minutes to run in isolation should probably isolated.

Command-Line Options

--maxSpecRunners

Type: Number Default: 8 The maximum number of webdrivers/browsers allowed to run in parallel.

--maxSpecFilesPerRunner

Type: Number Default: 'Spec files are divided among the available runners e.g. Math.ceil(options.specs.length / options.maxSpecRunners)` The maximum number of spec files to be included per runner. e.g. --maxSpecFilesPerRunner=1 is equivalent to using --isolateAllSpecs.

--isolateAllSpecs=true

Type: Boolean Default: false

Sets the maximum number of spec files to be included per runner to 1: e.g. --maxSpecFilesPerRunner=1 This is useful to find specs that cannot run in isolation. NOTE: Grab some coffee... this will take a while, I suggest that you pipe the output to a file. e.g. grunt test:faster --isolateAllSpecs > results.txt

options.keepalive

Type: Boolean Default: false

Prevents _SpecRunner.*.html files from being deleted and leaves the express server running. This is useful for debugging so that individual spec files can be run again manually

options.verbose

Type: Boolean Default: false

Specify the --verbose options to generate the full output from all subtask. In addition the --verbose options will cause the result summary to include the list of spec files for each runner.

options.burnin

Type: String

Specify the --burnin options to run a single spec file 100 times in parallel, the results of each run will be output to a table this can be useful for finding flaky test NOTE: you must specify the full relative path to the spec file e.g. grunt test:faster:firefox --burnin=rui/test/javascripts/rui/ui/notify/NotificationContainerSpec.js

Output

Includes:

• Generated runner files
• Spec files per runner
• Test per runner
• Time in seconds per runner
• Results per runner

Summary Row:

• Total spec files
• Total test run
• Total time in seconds (this value reflects what the total test time would have been if the runners were not run in parallel)

Total Execution Time: The actual elapsed time to run all test in parallel

Normal output will includes the spec files that had failures in each runner:

Verbose output includes the specs files that were included in each runner:

License

Copyright (c) 2015 Ronn McQueen Licensed under the MIT license.