python exposes an nt.environ for case insensitive environment behavior
that is native to windows; so it makes sense to use this instead of
os.environ to avoid enexpected behavior and failure.
further detail: https://bugs.python.org/issue28824
python exposes an nt.environ for case insensitive environment behavior
that is native to windows; so it makes sense to use this instead of
os.environ to avoid enexpected behavior and failure.
further detail: https://bugs.python.org/issue28824
Some hypervisors can handle several CPU architectures or have different
virtualization types. This is reflected in libvirt by the OS type (badly
named, indeed) and the arch value. Allow users to set them when creating
a VM using either virt.init or virt.running.
Signed-off-by: Cédric Bosdonnat <cbosdonnat@suse.com>
(cherry picked from commit 2463ebe5a8)
So far virt.running does nothing if the corresponding domain is already
defined. Use the new virt.update function to change the domain
configuration.
(cherry picked from commit 2a5f6ae5d6)
So far only the virt.running state is unit tested. Add tests for the
other states. Deprecated virt states have been purposedly left untested.
(cherry picked from commit fc75872fb6)
virt.running actually calls virt.init if the virtual machine doesn't
exist. Let the user define all the virt.init possible options in his
running states too.
(cherry picked from commit cb00a5f9b4)
Test virt.running when the domain needs to be edited and started. This
will prevent regressions when improving virt.running to handle the new
virt.init parameters.
(cherry picked from commit 495db345a5)
As mentioned in issue 47972, applying the virt.running state doesn't
report any error if the libvirt create call actually failed.
This commit introduces proper handling of the libvirt errors to let
users see the libvirt error in case of failure.
Also add test cases for virt.running to prevent regression.
(cherry picked from commit 451e7da55b)
The release tarball does not contain the `templates` directory.
Therefore `ExtendTestCase` will fail:
```
======================================================================
ERROR: test_run (unit.utils.test_extend.ExtendTestCase)
[CPU:0.0%|MEM:53.9%]
----------------------------------------------------------------------
Traceback (most recent call last):
File "tests/unit/utils/test_extend.py", line 40, in test_run
out = salt.utils.extend.run('test', 'test', 'this description', integration.CODE_DIR, False)
File "salt/utils/extend.py", line 242, in run
MODULE_OPTIONS = _fetch_templates(os.path.join(salt_dir, 'templates'))
File "salt/utils/extend.py", line 76, in _fetch_templates
for item in os.listdir(src):
FileNotFoundError: [Errno 2] No such file or directory: ' templates'
```
Signed-off-by: Benjamin Drung <benjamin.drung@cloud.ionos.com>
The release tarball does not contain `conf/cloud.profiles.d`,
`conf/cloud.providers.d`, and `conf/cloud.maps.d`. Therefore the test
cases will fail:
```
======================================================================
ERROR: test_conf_cloud_maps_d_files_are_commented (unit.test_config.SampleConfTest)
[CPU:0.0%|MEM:53.9%]
----------------------------------------------------------------------
Traceback (most recent call last):
File "tests/unit/test_config.py", line 236, in test_conf_cloud_maps_d_files_are_commented
cloud_sample_files = os.listdir(SAMPLE_CONF_DIR + 'cloud.maps.d/')
FileNotFoundError: [Errno 2] No such file or directory: 'conf/cloud.maps.d/'
======================================================================
ERROR: test_conf_cloud_profiles_d_files_are_commented (unit.test_config.SampleConfTest)
[CPU:0.0%|MEM:53.9%]
----------------------------------------------------------------------
Traceback (most recent call last):
File "tests/unit/test_config.py", line 200, in test_conf_cloud_profiles_d_files_are_commented
cloud_sample_files = os.listdir(SAMPLE_CONF_DIR + 'cloud.profiles.d/')
FileNotFoundError: [Errno 2] No such file or directory: 'conf/cloud.profiles.d/'
======================================================================
ERROR: test_conf_cloud_providers_d_files_are_commented (unit.test_config.SampleConfTest)
[CPU:0.0%|MEM:53.9%]
----------------------------------------------------------------------
Traceback (most recent call last):
File "tests/unit/test_config.py", line 218, in test_conf_cloud_providers_d_files_are_commented
cloud_sample_files = os.listdir(SAMPLE_CONF_DIR + 'cloud.providers.d/')
FileNotFoundError: [Errno 2] No such file or directory: 'conf/cloud.providers.d/'
```
Signed-off-by: Benjamin Drung <benjamin.drung@cloud.ionos.com>
Previously, when using `test=True`, `host.present` would return
`None` always. In Salt we return `True` when there are (or would be)
changes, `False` when there are no changes, and `None` when there
*would* have been changes.
This brings `host.present` in-line with that convention.
Rather than using `grains.items`, `test.arg` is less common, and
specifiying a value that isn't currently used anywhere else should
eliminate any chance of race condition for this test.
