NOTE: This also requires upgrading to tox-lsr 3.11.0
Ansible 2.19 will be released soon and has some changes which will
require fixes in system roles. This adds 2.19 to our testing matrix
on fedora 42 so that we can start addressing these issues.
python 3.13 is now being used on some platforms.
Using ansible-core 2.18 requires using py311 for pylint and other
python checkers.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
NOTE: This also requires upgrading to tox-lsr 3.11.0
Ansible 2.19 will be released soon and has some changes which will
require fixes in system roles. This adds 2.19 to our testing matrix
on fedora 42 so that we can start addressing these issues.
python 3.13 is now being used on some platforms.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
NOTE: This also requires upgrading to tox-lsr 3.10.0, and some
hacks to workaround a podman issue in ubuntu.
These tests run the role during a bootc container image build, deploy
the container into a QEMU VM, boot that, and validate the expected
configuration there. They run in two different tox environments, and
thus have to be run in two steps (preparation in buildah, validation in
QEMU). The preparation is expected to output a qcow2 image in
`tests/tmp/TESTNAME/qcow2/disk.qcow2`, i.e. the output structure of
<https://github.com/osbuild/bootc-image-builder>.
There are two possibilities:
* Have separate bootc end-to-end tests. These are tagged with
`tests::bootc-e2` and are skipped in the normal qemu-* scenarios.
They run as part of the container-* ones.
* Modify an existing test: These need to build a qcow2 image exactly
*once* (via calling `bootc-buildah-qcow.sh`) and skip setup/cleanup
and role invocations in validation mode, i.e. when
`__bootc_validation` is true.
In the container scenario, run the QEMU validation as a separate step in
the workflow.
See https://issues.redhat.com/browse/RHEL-88396
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
Add Fedora 42 to testing farm test matrix, drop Fedora 40
Use tox-lsr 3.9.0 for the `--lsr-report-errors-url` argument.
Add the argument `--lsr-report-errors-url DEFAULT` to the qemu test so that
the errors will be written to the output log. This uses the output callback
https://github.com/linux-system-roles/auto-maintenance/blob/main/callback_plugins/lsr_report_errors.py
Use the check_logs.py script
https://github.com/linux-system-roles/auto-maintenance/blob/main/check_logs.py
with the `--github-action-format` argument to format the errors
in a github action friendly manner.
Rename the log files `-FAIL.log` or `-SUCCESS.log` depending on status.
This is compatible with the way the testing farm log files are named, and
makes it easy to tell if a test passed or failed from the log file name.
Upload README.html as artifacts of the build_docs job for debugging
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
This will make the qemu/kvm tests be tested in either
ascending or descending ASCII order. This should give
us better test coverage of clean up scenarios which may
fail depending on the order of the previous tests.
Rename the qemu/kvm tests so that the statuses are shorter
and more intuitive.
Improve qemu/kvm test failure error reporting.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
These tests are problematic in github qemu tests, and that
functionality (scsi, anyway) in the testing farm integration
tests.
Yes, we should have a way to provide tags on a per-role basis . . .
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
tox-lsr 3.6.0 will guarantee order of qemu test execution, which should
help make tests reproducible and help debug test failures.
Improve qemu test logging - this will help debug the qemu test
failures.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
* You can ignore words inline by adding a comment like `# codespell:ignore word`.
* You can ignore words by adding them to the `.codespell_ignores` file.
* You can ignore files and directories by adding them with `skip = ` to the `.codespellrc` file.
Signed-off-by: Sergei Petrosian <spetrosi@redhat.com>
There is a new version of ansible-lint - v25.
Newer versions of ansible-lint require the collection requirements to be
installed so it can find the modules/plugins.
Enhance our ansible-lint ci job to provide the collection requirements,
including merging the runtime meta/collection-requirements.yml with
the testing tests/collection-requirements.yml.
This should somewhat mitigate the loss of ansible-plugin-scan.
We have to remove mock_modules that are actually present now.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
ansible-plugin-scan is broken due to lack of support for older versions
of python in ci.
One of the main reasons for using this scan is to check if the roles/tests
are using plugins that are not compatible with ansible 2.9. Since 2.9
is EOL, this is no longer necessary.
The other reason for using the scan is to check that the role/test
author has correctly listed dependencies in meta/collection-requirements.yml
and tests/collection-requirements.yml - that is - that the author has
correctly specified the dependencies for any plugins used that are
not built-in. This will mostly be caught in CI testing now.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
* Add "BusinessUnit": "system_roles" environment setting to tag our jobs in Testing farm
* Add tmt_plan_filter to run additional workflows besides general
* Allow more [citest bad] comment formats
* Get memory and supported platforms info from the PR ref
* Move LINUXSYSTEMROLES_USER to vars and use it everywhere in tft.yml
* Remove extra GITHUB_ORG definition
Signed-off-by: Sergei Petrosian <spetrosi@redhat.com>
* Remove plan from role dir, it's too complicated and long to run tests from tmt
* Use plan from linux-system-roles/tft-tests
* Move getting datetime to testing-farm job where it's used
* Add a workflow for running [citest_bad]
* Bump sclorg/testing-farm-as-github-action to v3
* Move concurrency group to first job
* Dump GitHub context
Signed-off-by: Sergei Petrosian <spetrosi@redhat.com>
* Add a preparation job to run on managed nodes
* Update instructions to run tests locally
* From tft workflow, remove condition for commenter author_association.
This would require us to add all trusted users, running workflow is not dangerous.
* Keep the plan minimal, only run prep steps required to install beakerlib
Signed-off-by: Sergei Petrosian <spetrosi@redhat.com>
This change is for running tests in Testing Farm CI. This is a replacement for
BaseOS CI that we are currently using. Running it Testing Farm gives us more
control.
It adds a workflow for running tests, and a plans directory containing a test
plan and a README-plans.md with some info.
Note that this workflow runs from the main branch. This means that changes to
the workflow must be merged to main, then pull requests will be able to run it.
This is because the workflow uses on: issue_comment context, this is a security
measure recommended by GitHub. It saves us from leaking organization secrets.
The functionality is WIP, so await future fixes and updates.
Signed-off-by: Sergei Petrosian <spetrosi@redhat.com>
the change made for
b4018c22f8
means that ansible-lint now needs an absolute path for the working directory
Go ahead and make ansible-test use absolute path too just in case they decide
to make the same change.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
The latest version of virtualenv does not support creating
python 2.7 virtualenvs. Change our CI tests to restrict the version
of virtualenv<20.22.0 and tox<4.15 for py27 environments
Move pylint, flake8, and black checks to the py310 environment
which is currently supported by ansible-core 2.17 and its related
checkers such as ansible-lint and ansible-test
pylint now uses ansible-core 2.17 and restricts the version of
pylint to 3.1.0 which is the version used by ansible-test 2.17
Remove `extends: default` for .yamllint.yml. The latest version
of ansible-lint will automatically incorporate local yamllint
settings unless there is an `extends:`.
The above changes require some fixes to the role code.
For more information, see
https://github.com/linux-system-roles/tox-lsr/pull/168
and
https://github.com/linux-system-roles/tox-lsr/pull/170
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
Upgrade ci tests to use tox-lsr 3.3.0
tox-lsr 3.3.0 uses ansible-test 2.17
Create the ansible-test ignore file for 2.17
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
This is fixed by tox-lsr 3.2.2 - all actions that use tox-lsr are updated to
3.2.2, not just the python unit tests, even though the fix is only related to
pytest. All roles are updated to use tox-lsr 3.2.2 for the sake of consistency
even if not affected by the pytest issue.
Something changed recently in the way github actions provisions systems which
means some of the directories are not readable by the python unit test actions.
In addition, the python unit tests were causing a lot of unnecessary directory
traversal doing collection/discovery of unit test files, because of using
`pytest -c /path/to/tox-lsr/pytest.ini` Unfortunately, with `pytest`, the
directory of the config file is the root directory for the tests and tests
discovery, and there is no way around this.
Therefore, the only solution is to copy the tox-lsr `pytest.ini` to the
`tests/unit` directory, which makes that the test root directory.
See also https://github.com/linux-system-roles/tox-lsr/pull/160
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
The old ansible-community ansible-lint is deprecated. There is a
new ansible-lint github action.
The latest Ansible repo gating tests run ansible-lint against
the collection format instead of against individual roles.
We have to convert the role to collection format before running
ansible-test.
This also requires tox-lsr 3.2.1 - bump other actions to use 3.2.1
Role developers can run this locally using
`tox -e collection,ansible-lint-collection`
See https://github.com/linux-system-roles/tox-lsr/pull/125
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
Use the `#!/usr/bin/env bash` shebang which is ansible-test friendly.
This means we can remove get_ostree_data.sh from the .sanity* files.
This also means we can remove the .sanity* files if we do not need
them otherwise. Fix other shell scripts to use the friendly shebang
and remove from the .sanity* files.
Rename `pth` to `path` in honor of nscott
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
This is primarily for the update to ansible-plugin-scan to
work with the upcoming ostree changes, but also includes
some minor fixes which affect ci.
3.1.0 was released but not used due to a bug fixed in 3.1.1
See full release notes for 3.1.0 and 3.1.1
https://github.com/linux-system-roles/tox-lsr/releases
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
This adds the dump_packages.py callback which will dump the
arguments to the `package` module (except for `state: absent`)
to the integration test run logs. The output looks like this:
`lsrpackages: pkg-a pkg-b ...`
We will have tooling which will scrape the logs to extract the
packages used at runtime and testing for all of the supported
combinations of distribution and version.
This also ensures the weekly-ci PR git commit message conforms
to commitlint.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
- Consistently generate badges for GH workflows in README RHELPLAN-146921
- Run markdownlint on all .md files
- Add custom-woke-action if not used already
- Use woke from linux-system-roles/lsr-woke-action
Signed-off-by: Sergei Petrosian <spetrosi@redhat.com>
- markdownlint runs against README.md to avoid any issues with
converting it to HTML
- test_converting_readme converts README.md > HTML and uploads this test
artifact to ensure that conversion works fine
- build_docs converts README.md > HTML and pushes the result to the
docs branch to publish dosc to GitHub pages site.
- Fix markdown issues in README.md
Signed-off-by: Sergei Petrosian <spetrosi@redhat.com>
We use the setup-python github action. This has dropped support for
python 2.7. Instead, just manually install the python2.7 package.
Signed-off-by: Rich Megginson <rmeggins@redhat.com>
