Writing and Running Barbican Tests

As a part of every code review that is submitted to the Barbican project there are a number of gating jobs which aid in the prevention of regression issues within Barbican. As a result, a Barbican developer should be familiar with running Barbican tests locally.

For your convenience we provide the ability to run all tests through the tox utility. If you are unfamiliar with tox please see refer to the tox documentation for assistance.

Unit Tests

Currently, we provide tox environments for Python 2.7. By default all available test environments within the tox configuration will execute when calling tox. If you want to run them independently, you can do so with the following command:

# Executes tests on Python 2.7
tox -e py27

Note

If you do not have the appropriate Python versions available, consider setting up PyEnv to install multiple versions of Python. See the documentation regarding Setting up a Barbican development environment for more information.

You can also setup breakpoints in the unit tests. This can be done by adding import pdb; pdb.set_trace() to the line of the unit test you want to examine, then running the following command:

# Executes tests on Python 2.7
tox -e debug

Note

For a list of pdb commands, please see: https://docs.python.org/2/library/pdb.html

Functional Tests

Unlike running unit tests, the functional tests require Barbican and Keystone services to be running in order to execute. For more information on setting up a Barbican development environment and using Keystone with Barbican, see our accompanying project documentation.

Once you have the appropriate services running and configured you can execute the functional tests through tox.

# Execute Barbican Functional Tests
tox -e functional

By default, the functional tox job will use nosetests to execute the functional tests. This is primarily due to nose being a very well known and common workflow among developers. It is important to note that the gating job will actually use testr instead of nosetests. If you discover issues while running your tests in the gate, then consider running testr or Devstack to more closely replicate the gating environment.

Remote Debugging

In order to be able to hit break-points on API calls, you must use remote debugging. This can be done by adding import rpdb; rpdb.set_trace() to the line of the API call you wish to test. For example, adding the breakpoint in def on_post in barbican.api.controllers.secrets.py will allow you to hit the breakpoint when a POST is done on the secrets URL.

..note:

After performing the ``POST`` the application will freeze. In order to use
``rpdb``, you must open up another terminal and run the following:

.. code-block:: bash

    # enter rpdb using telnet
    telnet localhost 4444

Once in rpdb, you can use the same commands as pdb, as seen here:
https://docs.python.org/2/library/pdb.html

Table Of Contents

Previous topic

Troubleshooting your Barbican Setup

Next topic

Plugin Developers Guide

This Page