How did you setup Capybara + Webdrivers on WSL for testing?

#help #discuss #testing #ruby

I recently started contributing to dev.to's GitHub repository by refactoring some .js files to meet the eslint specs. Some of the files that need to be refactored require multiple other ones to be refactored too at the same time in order to not break certain functionality.

In order to check for that I have been trying to setup the Acceptance testing as specified in the docs + with some online help to check whether all functionality remains as is or not. But I have been hitting various hurdles throughout where either the chrome binary wouldn't be found, post that after installation it posted Invalid Session id or sometimes it would work totally. It's a mystery for me currently.

Any help / pointer in the right direction would be much appreciated :D. Willing to provide any details required for the same!

Top comments (4)

Yash Dave • Sep 27 '19

So I managed to get tackle the chrome issue by using WSL's feature to access Windows binaries from Linux and calling the chrome installed to my windows. But the problem I am getting stuck at now is that it isn't able to perform any actions. It opens blank chrome console windows and just sits until it timesout with:

2.1) Failure/Error: visit "/#{user.username}/#{article.slug}/edit"

          Net::ReadTimeout:
            Net::ReadTimeout with #<TCPSocket:(closed)>



          # ./spec/system/articles/user_edits_an_article_spec.rb:20:in `block (2 levels) in <main>'
          # ./spec/rails_helper.rb:85:in `block (3 levels) in <top (required)>'
          # ./spec/rails_helper.rb:85:in `block (2 levels) in <top (required)>'
          # ./spec/support/initializers/rspec_retry.rb:8:in `block (2 levels) in <main>'

     2.2) Failure/Error: VCR.turned_off { ex.run }

          Net::ReadTimeout:
            Net::ReadTimeout with #<TCPSocket:(closed)>



          # ./spec/rails_helper.rb:85:in `block (3 levels) in <top (required)>'
          # ./spec/rails_helper.rb:85:in `block (2 levels) in <top (required)>'
          # ./spec/support/initializers/rspec_retry.rb:8:in `block (2 levels) in <main>'

Also a debug.log file is generated in the repo with content:

[0927/164130.672:ERROR:registration_protocol_win.cc(84)] TransactNamedPipe: The pipe has been ended. (0x6D)
[0927/164130.685:ERROR:file_io_win.cc(193)] LockFileEx: Incorrect function. (0x1)
[0927/164130.693:ERROR:file_io_win.cc(193)] LockFileEx: Incorrect function. (0x1)
[0927/164230.740:ERROR:file_io_win.cc(193)] LockFileEx: Incorrect function. (0x1)
[0927/164230.748:ERROR:file_io_win.cc(193)] LockFileEx: Incorrect function. (0x1)

Seems to be a firewall issue? But disabling windows firewall doesn't allow it through too.

Johnnathan Baxter • Apr 2 '21

Hi did you end up fixing this issue?

Yash Dave • Apr 3 '21 • Edited

Hey, the issues I was facing got resolved once WSL2 came out. With WSL2, it became possible to install chrome directly to my Linux distro without any problems. I followed the following steps for Ubuntu 18.04 with bash:

sudo apt install fonts-liberation libappindicator3-1 libasound2 libgbm1 libnspr4 libnss3 libxss1 xdg-utils
wget https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb
sudo dpkg -i google-chrome-stable_current_amd64.deb
echo 'alias chrome="google-chrome-stable"' >> ~/.bashrc
exec $SHELL

If you get an error while running dpkg, run sudo apt --fix-broken install after which run the dpkg command again. You can run the following command to test your install:

chrome --headless --disable-gpu --dump-dom https://www.chromestatus.com/

Mirzalazuardi Hermawan • Jun 11 '22

thanks , it works. this what I been searching for

DEV Community

How did you setup Capybara + Webdrivers on WSL for testing?

Top comments (4)

Read next

Writing Skills: Essential Fuel for Every Programmer 🥇👨‍💻

11 API Documentation Best Practices for CI/CD 2024

Which is best Next.JS or Nuxt.JS

Incorrect calculations: tan(x) and cot(x) for large values of x