Cucumber based blackbox testing framework for OpenShift

Getting started

More info you can find in Overview document.

Create local dev environment

1. Install git if not installed already, install ruby 2.4+ if not installed

2. git clone [email protected]:openshift/verification-tests.git

3. cd verification-tests

4. tools/install_os_deps.sh # need sudo; you may need to login again to terminal if ruby was installed via rvm during this phase (see below for RVM usage)

5. tools/hack_bundle.rb # normal user

6. Install OKD or OpenShift Container Platform client tools

7. Install the web driver if you desire to run tests which work with the Web UI. See Web Driver

Update local dev environment

bundle update

Test scenario execution

Check Configuring a BushSlicer v3 run

This is an example:

export BUSHSLICER_DEFAULT_ENVIRONMENT=ose
export OPENSHIFT_ENV_OSE_HOSTS=<host1>:etcd:master:node,<host2>:node,...
export OPENSHIFT_ENV_OSE_USER_MANAGER_USERS=user1:redhat,user2:redhat
# execute a whole feature file
bundle exec cucumber features/cli/create.feature
# or execute a single scenario
bundle exec cucumber features/cli/create.feature:5
# or execute the smoke tests
bundle exec cucumber --tags @smoke

For Origin environment, add the following environment variables:

export BUSHSLICER_DEFAULT_ENVIRONMENT=origin
export OPENSHIFT_ENV_ORIGIN_HOSTS=<host1>:etcd:master:node,<host2>:node,...
export OPENSHIFT_ENV_ORIGIN_USER_MANAGER_USERS=user1:redhat,user2:redhat

You can check which environments are defined in config/config.yaml.

You could also use the BUSHSLICER_CONFIG variable to override other configuration options. Add YAML/JSON into it and it will be merged with configuration file.

Debugging failures

If any step fails, you will fall into a pry shell where you can investigate what’s going on. This behavior depends on Cucumber profile that you enable during execution.

1. running steps using step 'Step name'

2. inspecting the @result variable

3. calling and methods in World

Running tests from outside this repository

You can put your private test scenarios under features/tierN directory. It is in .gitignore so that you can keep your test scenarios under the same file system tree without git messing them up.

Contributing

Please submit issues and pull requests to the project. You can discuss and ask questions on OpenShift mailing list.

RVM for dependency management

If you are planning on installing dependencies for verification Tests on your system, it is often a good idea to use some kind of Version Manager/Package manager, like RVM.

RVM is analogous to virtualenv in python and it stands for Ruby Version Manager.

You can install RVM and read more abou it on https://rvm.io/

Usually once RVM is installed, you would want to create/install a new rvm interpreter using:

When we install Ruby 2.7.1, it creates a local directory for your user with Ruby 2.7.1 under `~/.rvm/rubies/ruby-2.7.1`.
Once that is done, the Gemset creation is basically an attempt to have dependencies for various apps under same interpreter `2.7.1`. So in this case,
we are creating Gemset `verification-tests` for this repo dependencies. If you had multiple repos, you would use similar command as above to create a new Gemset under 2.7.1
and give it a name that you can recognize. Then for each respective app you may choose to use different gemset as `rvm use 2.7.1@verification-tests` or `rvm use 2.7.1@my-other-repo-or-app`

Feel free to read about interpreter and gemset in the Ruby docs or elsewhere on the internet.

Once ruby is installed, interpreter is created and gemset is present and activated, you can run `which ruby` to validate if the correct ruby is used and `ruby -v` to check version.
Following this, you may continue the steps as mentioned above, to run `bundle` commands to install gems listed in Gemfile to your Gemset.

Also if you are wondering what does `bundle` command do, search and read about `ruby bundler` on the internet.