5df748b2ed
During development of a new git commit, locally running a whole unit
or functional test suite to check every minor code change is
prohibitively expensive. For maximum developer productivity and
happiness, it's generally desirable to make the feedback loop of the
traditional red/green cycle as quick as possible.
So add run-tests-for-diff.sh and run-tests.py to the tools/
subdirectory, using a few tricks as explained below to help with this.
run-tests.py takes a list of files on STDIN, filters the list for
tests which can be run in the current tox virtualenv, and then runs
them with the correct stestr options.
run-tests-for-diff.sh is a simple wrapper around run-tests.py which
determines which tests to run using output from "git diff". This
allows running only the test files changed/added in the working tree:
tools/run-tests-for-diff.sh
or by a single commit:
tools/run-tests-for-diff.sh mybranch^!
or a range of commits, e.g. a branch containing a whole patch series
for a blueprint:
tools/run-tests-for-diff.sh gerrit/master..bp/my-blueprint
It supports the same "-HEAD" invocation syntax as flake8wrap.sh (as
used by the "fast8" tox environment):
tools/run-tests-for-diff.sh -HEAD
run-tests.py uses two tricks to make test runs as quick as possible:
1. It's (already) possible to speed up running of tests by
source'ing the "activate" file for the desired tox virtualenv,
e.g.
source .tox/py36/bin/activate
and then running stestr directly. This saves a few seconds by
skipping the overhead introduced by running tox.
2. When only one test file needs to be run, specifying the -n option
to stestr will skip the costly test discovery phase, saving
several more valuable seconds.
Future commits could build on top of this work, harnessing a framework
such as watchdog / watchmedo[0] or Guard[1] in order to automatically
run relevant tests every time your editor saves changes to a .py file.
[0] https://github.com/gorakhargosh/watchdog - Python-based
[1] https://guardgem.org - probably best in class, but Ruby-based so
maybe unacceptable for use within Nova.
Change-Id: I9a9bda5d29bbb8d8d77f769cd1abf7c42a18c36b
4.6 KiB
Test Strategy
A key part of the "four opens" is ensuring the OpenStack delivers well-tested and usable software. For more details see: http://docs.openstack.org/project-team-guide/introduction.html#the-four-opens
Experience has shown that untested features are frequently broken, in part due to the velocity of upstream changes. As we aim to ensure we keep all features working across upgrades, we must aim to test all features.
Reporting Test Coverage
For details on plans to report the current test coverage, refer to
/user/feature-classification.
Running tests and reporting results
Running tests locally
Please see https://opendev.org/openstack/nova/src/branch/master/HACKING.rst#running-tests
Voting in Gerrit
On every review in gerrit, check tests are run on very patch set, and are able to report a +1 or -1 vote. For more details, please see: http://docs.openstack.org/infra/manual/developers.html#automated-testing
Before merging any code, there is an integrate gate test queue, to ensure master is always passing all tests. For more details, please see: http://docs.openstack.org/infra/zuul/user/gating.html
Infra vs Third-Party
Tests that use fully open source components are generally run by the OpenStack Infra teams. Test setups that use non-open technology must be run outside of that infrastructure, but should still report their results upstream.
For more details, please see: http://docs.openstack.org/infra/system-config/third_party.html
Ad-hoc testing
It is particularly common for people to run ad-hoc tests on each released milestone, such as RC1, to stop regressions. While these efforts can help stabilize the release, as a community we have a much stronger preference for continuous integration testing. Partly this is because we encourage users to deploy master, and we generally have to assume that any upstream commit may already been deployed in production.
Types of tests
Unit tests
Unit tests help document and enforce the contract for each component. Without good unit test coverage it is hard to continue to quickly evolve the codebase. The correct level of unit test coverage is very subjective, and as such we are not aiming for a particular percentage of coverage, rather we are aiming for good coverage. Generally, every code change should have a related unit test: https://github.com/openstack/nova/blob/master/HACKING.rst#creating-unit-tests
Integration tests
Today, our integration tests involve running the Tempest test suite
on a variety of Nova deployment scenarios. The integration job setup is
defined in the .zuul.yaml file in the root of the nova
repository. Jobs are restricted by queue:
check: jobs in this queue automatically run on all proposed changes even with non-voting jobsgate: jobs in this queue automatically run on all approved changes (voting jobs only)experimental: jobs in this queue are non-voting and run on-demand by leaving a review comment on the change of "check experimental"
In addition, we have third parties running the tests on their preferred Nova deployment scenario.
Functional tests
Nova has a set of in-tree functional tests that focus on things that are out of scope for tempest testing and unit testing. Tempest tests run against a full live OpenStack deployment, generally deployed using devstack. At the other extreme, unit tests typically use mock to test a unit of code in isolation. Functional tests don't run an entire stack, they are isolated to nova code, and have no reliance on external services. They do have a WSGI app, nova services and a database, with minimal stubbing of nova internals.
Interoperability tests
The DefCore committee maintains a list that contains a subset of Tempest tests. These are used to verify if a particular Nova deployment's API responds as expected. For more details, see: https://github.com/openstack/defcore