RSpec + DatabaseCleaner + PhantomJS: database get randomly erased during example with js - ruby-on-rails

Question

I have a simple scenario in my rspec feature:
require 'spec_helper'
feature 'CLIENT views results page' do
context 'from welcome/index form' do
let!(:location) { FactoryGirl.create :location, name: 'town' }
before :each do
visit '/'
end
scenario 'successfully', js: true do
expect(Location.count).to eq 1
fill_in 'from_address', with: 'Some address'
fill_in 'to_address', with: 'Another address'
click_button 'Search'
expect(page).to have_content 'Some address → Another address'
end
end
end
spec_helper:
require 'rubygems'
ENV["RAILS_ENV"] ||= 'test'
require File.expand_path("../../config/environment", __FILE__)
require 'rspec/rails'
require 'factory_girl'
require 'database_cleaner'
require 'capybara/poltergeist'
FactoryGirl.reload
RSpec.configure do |config|
config.use_transactional_fixtures = false
config.infer_base_class_for_anonymous_controllers = true
config.order = "random"
config.tty = true
config.mock_with :rspec
config.filter_run focus: true
config.run_all_when_everything_filtered = true
config.treat_symbols_as_metadata_keys_with_true_values = true
config.before(:suite) do
DatabaseCleaner.clean_with(:truncation)
end
config.before(:each) do
DatabaseCleaner.strategy = :transaction
end
config.before(:each, js: true) do
DatabaseCleaner.strategy = :truncation
end
config.before(:each) do
DatabaseCleaner.start
end
config.after(:each) do
DatabaseCleaner.clean
end
end
Capybara.javascript_driver = :poltergeist
Capybara.asset_host = "http://localhost:3000"
Capybara.server_port = 9887 + ENV['TEST_ENV_NUMBER'].to_i
And my scenario fails on some random step (after expect(Location.count).to eq 1) due to the fact, that there are no Location records:
(byebug) Location.all
[]
I think it might be one of two cases:
Either DatabaseCleaner cleans database before an example finishes
Or there are actually two threads with two databases, one of which (the one that is used by the app) is really empty
Gem versions:
rspec (2.14.1)
rspec-core (2.14.8)
rspec-rails (2.14.0)
phantomjs (1.9.7.1)
rails (3.2.21)
database_cleaner (0.9.1)

Answer 1

I remember having issues with this as well. My config for DB cleaning looks like this:
config.before(:each) do
if Capybara.current_driver == :rack_test
DatabaseCleaner.strategy = :transaction
else
DatabaseCleaner.strategy = :truncation
end
DatabaseCleaner.start
end
config.after do
DatabaseCleaner.clean
end
Maybe it has something to do with the order your before blocks are called. So if DatabaseCleaner.start is called before you wanted strategy is set, you might see such results.

Answer 2

Managed to solve this issue by migrating to
config.use_transactional_fixtures = true
instead of using DatabaseCleaner. Some workarounds are suggested to make transactional cleaning work with javascript tests. I chose this one:
(adding somewhere in the beginning of spec/spec_helper)
class ActiveRecord::Base
mattr_accessor :shared_connection
##shared_connection = nil
def self.connection
##shared_connection || retrieve_connection
end
end
ActiveRecord::Base.shared_connection = ActiveRecord::Base.connection
cause it seemed to work faster, still leaving some spontaneously failing tests (might not be the problem related to database).