Investigating flakey tests — Run a test loads in rspec by wrapping in a context

When trying to reproduce flakiness I used to just manually run the test x times from the command line and count how many times it had failed and what the errors were.

Doing it in the code saved me a bunch of time.

# original
it "is tells us it happened" do
  user = create(:user)
  visit "/"
  click_link "make it happen"

  expect(page).to have_content "it happened"
end


# run it loads
context "when testing for flakiness" do
  20.times do
    it "is tells us it happened" do
      user = create(:user)
      visit "/"
      click_link "make it happen"

      expect(page).to have_content "it happened"
    end
  end
end