GOV.UK Publishing End-to-end Tests
A suite of end-to-end publishing tests that allow us to test functionality across applications and services. The idea is that we test the sequence of actions and movement of data throughout the system in a 'real world' context, rather than stubbing services and making assumptions about responses. The tests are browser tests (written in RSpec, using Capybara) that mimic the behaviour of content editors.
Tests are written against a variety of publishing applications to see which apps are tested check the contents of the spec directory. To view the details of all the apps involved check docker-compose.yml.
How to run the tests
If this is your first time running the E2E project make sure you have installed Docker and run:
$ bundle install
Build and run the test suite with:
$ make -j4
If it has been some time since you last worked on the E2E project it is recommended to instead run:
$ make clean $ make -j4
This will remove all your local apps, clone them, check them out to the latest deployed-to-production branch. This has been a problem in the past due to stale branches being checked out and either being unbuildable or causing false test failures.
Running this command executes the following targets in order, which you can
choose to run separately to speed up development:
For example, to run only the tests for the specialist publisher, you need only do:
$ make -j4 clone $ make pull build start test-specialist-publisher stop
If you need to run the tests against a branch of an application other than deployed-to-production you need to explicitly build it as below:
$ make -j4 clone pull $ docker-compose build publisher $ make start test-publisher stop
When making changes to an application you will need to rebuild the image before the new version will be used.
$ docker-compose build publisher
When you have finished testing against your branch version and want to switch back
to the deployed-to-production version you will need to untag the built image before
you can re-pull. The
clean_docker make recipe will untag all locally built images.
$ make clean_docker $ make pull
See docs/docker.md for more information configuring/troubleshooting docker.
It is possible the run this suite without Docker however you will have to
set-up and configure each application in the stack and have them available
*.dev.gov.uk. The tests can then be run with:
bundle exec rspec
Failing Jenkins builds
We have detailed guidance to help with debugging the reason for a Jenkins failure inside the docs folder.
Additionally, if you are making a change to an E2E tested application which requires a change one of these tests, we have detailed guidance on how to achieve that.
Dealing with a flaky test
A key aim of these tests is to be as reliable as possible, however this can be difficult - particularly early in a test's introduction - therefore we have a strategy to deal with flaky tests.
There is information on the coding standards, how to add tests and how to add applications to this project in the contributing guidelines.
WEBrick server seems to stop responding
We had a nasty bug with Router API where the web server seemed to stop responding to any requests. This issue was experienced once the application was upgraded to Rails 5.1 and Mongoid 6.1. The symptom was that any requests to the server seemed to hang and never respond. The way this was resolved was to switch from using WEBrick on Router API and instead use unicorn server.
Page caching issue
We encountered an issue with Poltergeist caching a page that was used in two different tests. This would result in whichever test ran second failing as the excpected text would not show. We resolved this by clearing the Poltergeist page cache before each test.
This issue could still potentially arise if a test visited the same page multiple times in the same test and would likely need a similar solution to be included as part of the test.