cinder/doc/source/contributor/testing.rst
Andreas Jaeger dff0f7b2c1 Docs: Jenkins is dead, long live Zuul
Cinder does not use Jenkins for quite some time anymore, update the
information for Zuul.

Fix also an aligment problem and a formatting bug (_x_ is
not RST markup) in file i18n.rst.

Change-Id: Icaa26764041106ad95eb69b7a94400429e025bd9
2019-08-13 21:06:27 +02:00

6.6 KiB

Testing

Cinder contains a few different test suites in the cinder/tests/ directory. The different test suites are Unit Tests, Functional Tests, and Tempest Tests.

Test Types

Unit Tests

Unit tests are tests for individual methods, with at most a small handful of modules involved. Mock should be used to remove any external dependencies.

All significant code changes should have unit test coverage validating the code happy path and any failure paths.

Any proposed code change will be automatically rejected by the OpenDev Zuul project gating system1 if the change causes unit test failures.

Functional Tests

Functional tests validate a code path within Cinder. These tests should validate the interaction of various modules within the project to verify the code is logically correct.

Functional tests run with a database present and may start Cinder services to accept requests. These tests should not need to access an other OpenStack non-Cinder services.

Tempest Tests

The tempest tests in the Cinder tree validate the operational correctness between Cinder and external components such as Nova, Glance, etc. These are integration tests driven via public APIs to verify actual end user usage scenarios.

Running the tests

There are a number of ways to run tests currently, and there's a combination of frameworks used depending on what commands you use. The preferred method is to use tox, which calls ostestr via the tox.ini file.

Unit Tests

To run all unit tests simply run:

tox

This will create a virtual environment, load all the packages from test-requirements.txt and run all unit tests as well as run flake8 and hacking checks against the code.

You may run individual test targets, for example only py27 tests, by running:

tox -e py27

Note that you can inspect the tox.ini file to get more details on the available options and what the test run does by default.

Functional Tests

To run all functional tests, run:

tox -e functional

Tempest Tests

Tempest tests in the Cinder tree are "plugged in" to the normal tempest test execution. To ensure the Cinder tests are picked up when running tempest, run:

cd /opt/stack/tempest
tox -e all-plugin

More information about tempest can be found in the Tempest Documentation.

Database Setup

Some unit and functional tests will use a local database. You can use tools/test-setup.sh to set up your local system the same way as it's setup in the CI environment.

Running a subset of tests using tox

One common activity is to just run a single test, you can do this with tox simply by specifying to just run py27 or py35 tests against a single test:

tox -epy27 -- cinder.tests.unit.volume.test_availability_zone.AvailabilityZoneTestCase.test_list_availability_zones_cached

Or all tests in the test_volume.py file:

tox -epy27 -- cinder.tests.unit.volume.test_volume

You may also use regular expressions to run any matching tests:

tox -epy27 -- test_volume

For more information on these options and details about stestr, please see the stestr documentation.

Gotchas

Running Tests from Shared Folders

If you are running the unit tests from a shared folder, you may see tests start to fail or stop completely as a result of Python lockfile issues. You can get around this by manually setting or updating the following line in cinder/tests/conf_fixture.py:

CONF['lock_path'].SetDefault('/tmp')

Note that you may use any location (not just /tmp!) as long as it is not a shared folder.

Running py35 tests

You will need to install python3-dev in order to get py35 tests to run. If you do not have this, you will get the following:

netifaces.c:1:20: fatal error: Python.h: No such file or directory
    #include <Python.h>
            ^
    compilation terminated.
    error: command 'x86_64-linux-gnu-gcc' failed with exit status 1

    ----------------------------------------
    <snip>
ERROR: could not install deps [-r/opt/stack/cinder/test-requirements.txt,
    oslo.versionedobjects[fixtures]]; v = InvocationError('/opt/stack/cinder/
    .tox/py35/bin/pip install -r/opt/stack/cinder/test-requirements.txt
    oslo.versionedobjects[fixtures] (see /opt/stack/cinder/.tox/py35/log/py35-1.log)', 1)
_______________________________________________________________ summary _______________________________________________________________
ERROR:   py35: could not install deps [-r/opt/stack/cinder/test-requirements.txt,
    oslo.versionedobjects[fixtures]]; v = InvocationError('/opt/stack/cinder/
    .tox/py35/bin/pip install -r/opt/stack/cinder/test-requirements.txt
    oslo.versionedobjects[fixtures] (see /opt/stack/cinder/.tox/py35/log/py35-1.log)', 1)

To Fix:

  • On Ubuntu/Debian:

    sudo apt-get install python3-dev
  • On Fedora 21/RHEL7/CentOS7:

    sudo yum install python3-devel
  • On Fedora 22 and higher:

    sudo dnf install python3-devel

Assertion types in unit tests

In general, it is best to use the most specific assertion possible in a unit test, to have the strongest validation of code behavior.

For example:

self.assertEqual("in-use", volume.status)

is preferred over

self.assertIsNotNone(volume.status)

or

Test methods that implement comparison checks are also generally preferred over writing code into assertEqual() or assertTrue().

self.assertGreater(2, volume.size)

is preferred over

self.assertTrue(2 > volume.size)

However, assertFalse() behavior is not obvious in this regard. Since None evaluates to False in Python, the following check will pass when x is False or None.

self.assertFalse(x)

Therefore, it is preferable to use:

self.assertEqual(x, False)

Footnotes

Debugging

Debugging unit tests

It is possible to attach a debugger to unit tests.

First, modify the test you want to debug by adding the following to the test code itself:

import pdb
pdb.set_trace()

Then run the unit test with pdb enabled:

source .tox/py36/bin/activate

stestr run -n cinder.tests.unit.test_volume_utils

# Or to get a list of tests to run

stestr list test_volume_utils > tests_to_run.txt
stestr run -n --load-list tests_to_run.txt

  1. See zuul.↩︎