Some integration test targets have dependencies on files outside
the `test/integration/targets/` directory tree. Changes to these
dependencies can result in unexpected test failures since they do
not trigger integration tests which depend on them.
* Log dependencies at verbosity level 4.
This makes it easier to debug target dependency issues.
* Scan symlinks for target dependencies.
Some test targets use symlinks to files in other test targets.
These dependencies were previously undetected. This could result in
changes made to dependencies without triggering the dependent tests.
* Track missing target deps with `needs/target/*`.
Some existing test targets have untracked dependencies on other
test targets. This can result in changes to those dependencies
not triggering their dependent tests, resulting in test failures
after a PR is merged.
This PR adds the appropriate `needs/target/*` aliases to track
those dependencies, along with appropriate processing in
ansible-test to handle the new aliases.
* Scan meta dependencies in script targets.
Script targets are often former role targets which were converted
to allow custom invocations of ansible-playbook. These targets still
have their meta dependencies, but they were not being detected.
This could result in changes to dependencies not triggering the
targets which depend on them.
Previously, the following dependencies:
A used by B
B used by C
Would have been converted to:
A used by C
B used by C
Intead of being expanded to:
A used by B
A used by C
B used by C
This change preserves the existing dependency when expanding it.
* Move var_blending test inventory into test.
* Remove Amazon specific inventory entry for tests.
* Remove Azure specific inventory entry for tests.
* Move var_precedence test inventory into test.
* Move unicode test inventory into test.
* Remove unused inventory entry.
* Move gathering_facts test inventory into test.
* Move delegate_to test inventory into test.
* Clean up inventory for binary_modules test.
* Clean up integration test inventory.
* Cloudscale integration test setup
CloudProvider and CloudEnvironment classes for Cloudscale integration
tests. This also contains a cloudscale_common role with common
variables for all tests.
* cloudscale_volume module
New cloud module to manage volumes on the cloudscale.ch IaaS service.
It is currently supported only with the `--remote` option.
This makes it easier to troubleshoot new instances which are not
yet supported by the setup scripts used by ansible-test.
* Support skip of platforms by version in tests.
Previously a remote platform could be skipped completely using the alias:
`skip/{platform}` such as `skip/rhel`
Now a specific platform version can be skipped using the alias:
`skip/{platform}{version}` such as `skip/rhel7.6`
This feature is available for platforms specified with the `--remote` option.
* Add skip by version to the docs.
* Provide Kubernetes resource validation to k8s module
Use kubernetes-validate to validate Kubernetes resource
definitions against the published schema
* Additional tests for kubernetes-validate
* Improve k8s error messages on exceptions
Parse the response body for the message rather than returning
a JSON blob
If we've validated and there are warnings, return those too - they
can be more helpful
```
"msg": "Failed to patch object: {\"kind\":\"Status\",\"apiVersion\":\"v1\",\"metadata\":{},
\"status\":\"Failure\",\"message\":\"[pos 334]: json: decNum: got first char 'h'\",\"code\":500}\n",
```
vs
```
"msg": "Failed to patch object: [pos 334]: json: decNum: got first char 'h'\nresource
validation error at spec.replicas: 'hello' is not of type u'integer'",
```
* Update versions used
In particular openshift/origin:3.9.0
* Add changelog for k8s validate change
Track the interpreter for each copy of the injector by the interpreter
path instead of the interpreter version. This avoids the possibility
of mixing different interpreters with the same version.
Inject a symlink to the correct python into the copied injector
directory instead of altering the shebang of the injector. This
has the side-effect of also intercepting `python` for integration
tests which simplifies cases where it needs to be directly invoked
without collecting code coverage.
* Add support for POST-as-GET if GET fails with 405.
* Bumping ACME test container version to 1.4. This includes letsencrypt/pebble#162 and letsencrypt/pebble#168.
* Also use POST-as-GET for account data retrival.
This is not yet supported by any ACME server (see letsencrypt/pebble#171),
so we fall back to a regular empty update if a 'malformedRequest' error is
returned.
* Using newest ACME test container image.
Includes letsencrypt/pebble#171 and letsencrypt/pebble#172, which make Pebble behave closer to the current specs.
* Remove workaround for old Pebble version.
* Add changelog entry.
* First try POST-as-GET, then fall back to unauthenticated GET.
Trying to get ansible-test working on my fedora-28 system, I noticed I
was getting invalid keys from paramiko. It looks like this is because
ssh-keygen is now defaulting to RFC4716 format for private / public
keys.
For now, we can still use PEM based SSH keys, but the long term fix here
is to report a bug to paramiko and support RFC4716 for rsa keys.
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
* Move ansible.compat.tests to test/units/compat/.
* Fix unit test references to ansible.compat.tests.
* Move builtins compat to separate file.
* Fix classification of test/units/compat/ dir.
* ansible-test: add skip/windows/... alias to skip tests on specific Windows versions
* show what tests were skipped
* changes to logic to only skip if all Windows targets are set to skip
* codestyle improvements
* change warning message based on review
* check args type before running the Windows path
* Add symlinks sanity test.
* Replace legacy test symlinks with actual content.
* Remove dir symlink from template_jinja2_latest.
* Update import test to use generated library dir.
* Fix copy test symlink setup.
* Add unified diff output to environment validation.
This makes it easier to see where the environment changed.
* Compare Python interpreters by version to pip shebangs.
This helps expose cases where pip executables use a different
Python interpreter than is expected.
* Query `pip.__version__` instead of using `pip --version`.
This is a much faster way to query the pip version. It also more
closely matches how we invoke pip within ansible-test.
* Remove redundant environment scan between tests.
This reuses the environment scan from the end of the previous test
as the basis for comparison during the next test.
* win_exec: refactor PS exec runner
* more changes for PSCore compatibility
* made some changes based on the recent review
* split up module exec scripts for smaller payload
* removed C# module support to focus on just error msg improvement
* cleaned up c# test classifier code
* Fix ansible-test smoke tests across groups.
* Fix ansible-test list arg defaults.
* Fix ansible-test require and exclude delegation.
* Fix detection of Windows specific changes.
* Add minimal Windows testing for Python 3.7.
This removes the old name based version detection behavior and
uses versions defined in the docker completion file instead, as
the new containers do not follow the old naming scheme.
This prevents tests from loading modules outside the source tree,
which could result in testing the wrong module if a system-wide
install is present, or custom modules exist.
This can be used to run Python scripts from the repository with the
correct interpreter and allow collection of code coverage.
Useful for testing contrib inventory scripts.
* Make merge_type a list and apply merge_type in order
Allow use case of preferring strategic-merge and failing
back to merge, or just preferring a different merge type
* Improve k8s module test coverage
* new nios module support
* new nios module support
* new nios module support
* new nios module support
* new nios module support
* new nios module support
* new nios module support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* new nios module integration test support
* test/integration/targets/nios_naptr_record/tasks/nios_naptr_record_idempotence.yml
new nios module integration test support
* fix pep8 error
* fix pep8 error
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* adding newline at end
* Using ACME test container for acme_account integration test.
* Removing dependency on setup_openssl. Waiting for controller and Pebble.
* More tinkering.
* Reducing number of tries.
* One more try.
* Another try.
* Added acme_certificate tests.
* Removed double key.
* Added tests for acme_certificate_revoke.
* Making task names more meaningful (during certificate generation).
* Using newer test container which integrates letsencrypt/pebble#137. Adding test for revoking certificate by its private key.
* Using new version of Pebble which limits the random auth delay.
* Simplifying certificates for revocation tests.
* Reworking acme_certificate tests (there are now more, but they are faster).
* Test whether account_key_content works.
* Preparing TLS-ALPN-01 support.
* Using official Ansible image of testing container on quay.io.
* Bumping version.
* Bumping version of test container to 1.1.0.
* Adjusting to new CI group names.
* Pass ACME simulator IP as playbook variable.
* Let test plugin wait for controller and CA endpoints to become active.
* Refactor common setup parts of tests to setup_acme.
* _ -> dummy
* Moving common obtain-cert.yml to setup_acme.
Now that we don't need to worry about python-2.4 and 2.5, we can make
some improvements to the way AnsiballZ handles modules.
* Change AnsiballZ wrapper to use import to invoke the module
We need the module to think of itself as a script because it could be
coded as:
main()
or as:
if __name__ == '__main__':
main()
Or even as:
if __name__ == '__main__':
random_function_name()
A script will invoke all of those. Prior to this change, we invoked
a second Python interpreter on the module so that it really was
a script. However, this means that we have to run python twice (once
for the AnsiballZ wrapper and once for the module). This change makes
the module think that it is a script (because __name__ in the module ==
'__main__') but it's actually being invoked by us importing the module
code.
There's three ways we've come up to do this.
* The most elegant is to use zipimporter and tell the import mechanism
that the module being loaded is __main__:
* 5959f11c9d/lib/ansible/executor/module_common.py (L175)
* zipimporter is nice because we do not have to extract the module from
the zip file and save it to the disk when we do that. The import
machinery does it all for us.
* The drawback is that modules do not have a __file__ which points
to a real file when they do this. Modules could be using __file__
to for a variety of reasons, most of those probably have
replacements (the most common one is to find a writable directory
for temporary files. AnsibleModule.tmpdir should be used instead)
We can monkeypatch __file__ in fom AnsibleModule initialization
but that's kind of gross. There's no way I can see to do this
from the wrapper.
* Next, there's imp.load_module():
* https://github.com/abadger/ansible/blob/340edf7489/lib/ansible/executor/module_common.py#L151
* imp has the nice property of allowing us to set __name__ to
__main__ without changing the name of the file itself
* We also don't have to do anything special to set __file__ for
backwards compatibility (although the reason for that is the
drawback):
* Its drawback is that it requires the file to exist on disk so we
have to explicitly extract it from the zipfile and save it to
a temporary file
* The last choice is to use exec to execute the module:
* https://github.com/abadger/ansible/blob/f47a4ccc76/lib/ansible/executor/module_common.py#L175
* The code we would have to maintain for this looks pretty clean.
In the wrapper we create a ModuleType, set __file__ on it, read
the module's contents in from the zip file and then exec it.
* Drawbacks: We still have to explicitly extract the file's contents
from the zip archive instead of letting python's import mechanism
handle it.
* Exec also has hidden performance issues and breaks certain
assumptions that modules could be making about their own code:
http://lucumr.pocoo.org/2011/2/1/exec-in-python/
Our plan is to use imp.load_module() for now, deprecate the use of
__file__ in modules, and switch to zipimport once the deprecation
period for __file__ is over (without monkeypatching a fake __file__ in
via AnsibleModule).
* Rename the name of the AnsiBallZ wrapped module
This makes it obvious that the wrapped module isn't the module file that
we distribute. It's part of trying to mitigate the fact that the module
is now named __main)).py in tracebacks.
* Shield all wrapper symbols inside of a function
With the new import code, all symbols in the wrapper become visible in
the module. To mitigate the chance of collisions, move most symbols
into a toplevel function. The only symbols left in the global namespace
are now _ANSIBALLZ_WRAPPER and _ansiballz_main.
revised porting guide entry
Integrate code coverage collection into AnsiballZ.
ci_coverage
ci_complete
* Update Shippable integration test groups.
* Update integration test group aliases.
* Rebalance AWS and Azure tests with extra group.
* Rebalance Windows tests with another group.
* leave vmware_guest_powerstate tests enabled, but revert changes from: 66743f33
* leave VM poweroff tests enabled, but revert changes from: 87d6bdaf
* bumped 'vcenter-test-container' version to '1.3.0'
* updated test task names based on PR feedback
* add aks module and integration tests
* linting
* update tests
* sanity check
* make some changes to AKS module
* make integration test work
* add fact
* add resource_group name
* add fact test
* fix test
* fix test
* linting
* changed line endings for facts
* output kubeconfig
* Update azure_rm_aks.py
* update integration test aliases
* update aliases
* add cloud_environment and auth_source to args
* Fix comments from Jborean93 (#3)
* update
* fix
* fix
* fix
* fix
* update doc
* fix
Previously the test framework (DCI, Zuul) were installing the various
dependencies, this meant the list of what was required was duplicated.
Having everything defined in ansible-test makes it easier for people to
run tests locally.
Also this allows the test to work correctly on Python 2 & Python 3
- Works with the --remote option.
- Can be disabled with the --disable-httptester option.
- Change image with the --httptester option.
- Only load and run httptester for targets that require it.
