tests.support.helpers passes the single-asterisk tuple of args as
individual arguments to runtest_helpers.modules_available. This breaks
tests which pass more than one argument to the requires_salt_modules
decorator.
This has been fixed by making runtests_helpers.modules_available accept
an arbitrary number of args via using a single-asterisk.
If this is has any unicode characters in it, it won't load on systems that do
not default to a unicode locale.
find . -type f | while read line; do ret=$(file $line); if [[ $ret == *UTF-8* && $line == *.py ]]; then echo $line; fi; done
The above will list all files that have unicode characters in it and won't load
with locale set to C or POSIX
We need to make the comparison in the file to `7` instead of `'7'`,
otherwise the state doesn't execute correctly and the test_pkgrepo
test will fail because the state didn't run.
Fixes#38683
When two states combine the `require`, `failhard`, and `order` options,
the `order` option should be ignored because `require` is present. Then
the `failhard` option should make the state run fail.
The check for "failhard" in the `check_failhard` function in the state
compiler was too broad. We want "failhard" to be true AND tag to be in
the `running` dict. Without the parens around the first OR statement,
we were bypassing the AND requirement because failhard was found.
Since the second state's tag was not found in the running dict, the state
run stacktraces on a KeyError.
This does the following:
- Splits states for container/volume/image/network management into four
separate state modules.
- Preserves backward compatibility by making ``docker.image_present``
invoke ``docker_image.present``, etc.
- Changes how Salt detects that a container needs to be replaced.
Instead of comparing each passed argument to the named container's
configuration, it creates a temporary container, and compares that
container to the named container. If the two differ, then the older
container is removed, and the new one is renamed and started, becoming
the named container.
- Removes the unit tests for container management and replaces them with
integration tests.
- Adds unit tests for the new salt.utils.docker
This commit also modifies the master config for the test suite to use
this new option, so that a missing GitPython/Pygit2 will not crash the
test suite's master. This should make the test suite run smoother by
only negatively impacting those tests which require the pillar data
provided by git_pillar, when a valid provider is not available.
This uses a function in the runtests_helpers custom module to perform
all the logic, and only returns what failed the test. This saves us from
having to return the entire contents of sys.doc (as well as log all of
the function calls), and also removes the need to run sys.doc in batches
to get around the "max message size" issue.
Client was only setting success to false if the function called raised
an exception. This commit changes it to verify the content of return
value with check_state_result
Client was only setting success to false if the function called raised
an exception. This commit changes it to verify the content of return
value with check_state_result
* Fix typo in profile example ('private_key' listed twice)
* Reflect Joyent's current new naming convention for VM sizes
* Update docs with modern images that are officially supported by Joyent
* Refresh example output of --list-sizes and --list-images
* Fix typo in profile example ('private_key' listed twice)
* Reflect Joyent's current new naming convention for VM sizes
* Update docs with modern images that are officially supported by Joyent
* Refresh example output of --list-sizes and --list-images
When running the tests with the tcp transport, we are not as forgiving
with the minion connection process as we are in ZMQ. In ZMQ, we attempt
to connect to the master. If it isn't up yet, we wait and try again. In
TCP, we try to connect to the master once, realize it's not up (because
the master process takes longer to spin up than the minions) and crash
and bail out.
This just gives the master a little more time to come up by having the
minions try to connect a couple more times.
The test that runs these states is testing for behavior that was
obsoleted by virtualenv 13.0. Ensure that we have older virtualenv
available, and then create a venv with that older version. Use the
2nd virtualenv to attempt the "weird" install.
The test that runs these states is testing for behavior that was
obsoleted by virtualenv 13.0. Ensure that we have older virtualenv
available, and then create a venv with that older version. Use the
2nd virtualenv to attempt the "weird" install.
* OSX is a bit more restrictive on getting the process information
* OSX says the sockets aren't connected, which is weird, but...
* Close the sockets on OSX
* Refactored LAN and NIC provisioning to use a more efficient composite server request.
* Added support for new INTEL_XEON and AMD_OPTERON CPU family types.
* Allow additional data volumes can be created and attached to servers.
* Added public and private network firewall rules.
* Added preliminary support for loadbalancers CLI functions.
* Corrected ProfitBricks name topic index.
* Updated ProfitBricks provider documentation.
* Updated provider integration test configs.
* PEP8 clean up.
* Add integration test for #34945
* file.recurse: Do not convert octal mode string to int
When we run file.makedirs_perms to create the dest directory, we pass
through the mode to file.check_perms. However, file.check_perms expects
an octal string, not an int. This causes the initial directory to be
chmod'ed to the wrong mode. When there are files in the source
directory, file.recurse will invoke the file.directory state to manage
files/dirs in that directory, and this ends up correcting the mode as we
simply pass the dir_mode to it. However, when there are only directories
in the source directory, this never happens and the incorrect mode
remains on the destination directory.
Fixes#34945.
This got removed during the attempt to make the tests run on Windows.
I added them everywhere even though that's not strictly necessary.
Without these set on the master and sub_master, Windows will throw
erorrs because the networking stack will decide that anything that's
bound to 0.0.0.0 represents an extestential threat to any socket that
attempts to *connect* to one of those ports. (For...reasons.)
* Remove redundant usage of destructiveTest decorator
* Remove unnecessary SLS file
This simple state can be run using run_state
* Explicitly order tests
* Use run_state instead of invoking state.sls
* Ensure only one refresh_db done during pkg integration tests
* Add pkg.latest test for package without an epoch
* Optimize pkg.latest_version usage in pkg integration tests
* Clarify outcome when only_upgrade is used
* Add pkg.latest test with only_upgrade=True
* better support for named images, locations, sizes
* better support for named images, locations, sizes
* Tests for Vultr driver
* Add sleep so test instance will be old enough to delete as part of the test
* Lint.
* Add run_on_start docs to schedule.rst (#32958)
Fixes#22580
* Backport #33021 manually to 2015.5 (#33044)
* Saltfile with pillar tests (#33045)
* add file.managed with pillar data tests
* do not require git for other tests
* Whitespace...
* modules.virtualenv_mod: use correct pip bootstrap url
* modules.pip: raise error on mirrors arg
* states.pip: run mirrors test on < 7.0.0
* update pip integration test states to not use mirrors
* modules.pip: run mirrors tests on pip < 7.0.0
* fix sorting by latest version when called with an attribute
* remove reference to master_alive_check
* Fixes saltstack/salt#28262
* Resolve memory leak in authentication
* outputter virt_list does not exist anymore
* Update proxmox documentation
* Fix documentation on boto_asg and boto_elb modules and states
* modules.win_timezone: don't list all zones in debug log
* Correcty index glusterfs bricks
Fixes issue #32311
* Cleaner deprecation process with decorators
* Add deprecation decorator scaffold
* Capture type error and unhandled exceptions while function calls
* Aware of the current and future version of deprecation
* Implement initially is_deprecated decorator
* Add an alias for the capitalization
* Fix capitalization easier way
* Remove an extra line
* Add successor name to the deprecation decorator.
* Granulate logging and error messages.
* Implement function swapper
* Raise later the caught exception
* Clarify exception message
* Save function original name
* Remove an extra line
* Hide an alternative hidden function name in the error message, preserving the error itself
* Rename variable as private
* Add a method to detect if a function is using its previous version
* Message to the log and/or raise an exception accordingly to the status of used function
* Log an error along with the exception
* Add internal method documentation
* Add documentation and usage process for decorator "is_deprecated"
* Add documentation and process usage for the decorator "with_deprecated"
* Hide private method name
* Fix PEP8, re-word the error message
* Deprecate basic uptime function
* Add initial decorator unit test
* Rename old/new functions, mock versions
* Move frequent data to the test setup
* Add logging on EOL exception
* Rename and document high to low version test on is_deprecated
* Implement a test on low to high version of is_deprecated decorator
* Add a correction to the test description
* Remove a dead code
* Implement a test for high to low version on is_deprecated, using with_successor param
* Correct typso adn mistaeks
* Implement high to low version with successor param on is_deprecated
* Setup a virtual name for the module
* Implement test for with_deprecated should raise an exception if same deprecated function not found
* Implement test for with_deprecated an old function is picked up if configured
* Correct test description purpose
* Implement test with_deprecated when no deprecation is requested
* Add logging test to the configured deprecation request
* Add logging testing when deprecated version wasn't requested
* Implement test EOL for with_deprecated decorator
* Correct test explanation
* Rename the test
* Implement with_deprecated no EOL, deprecated other function name
* Implement with_deprecated, deprecated other function name, EOL reached
* Add test description for the with_deprecated + with_name + EOL
* Fix confusing test names
* Add logging test to the is_deprecated decorator when function as not found.
* Add more test point to each test, remove empty lines
* Bugfix: at certain conditions a wrong alias name is reported to the log
* Fix a typo in a comment
* Add test for the logging
* Disable a pylint: None will _never_ be raised
* Fix test for the deprecated "status.uptime" version
* Bugfix: Do not yank raised exceptions
* Remove unnecessary decorator
* Add test for the new uptime
* Add test for the new uptime fails when /proc/uptime does not exists
* Rename old test case
* Skip test for the UTC time, unless freeze time is used.
* Fix pylint
* Fix documentation
* Bugfix: proxy-pass the docstring of the decorated function
* Lint fix
* Fixessaltstack/salt#28262 for 2015.5 branch
* Update master config docs
* Improve git_pillar documentation/logging
* Add note about different behavior of top file in git_pillar
* Make log entry for a missing pillar SLS file more accurate for git_pillar
* FreeBSD supports packages in format java/openjdk7 so the prior commit broke that functionality. Check freebsd/pkg#1409 for more info.
* FreeBSD supports packages in format java/openjdk7 so the prior commit broke that functionality. Check freebsd/pkg#1409 for more info.
* Update glusterfs_test to be inline with #32312
* Fix salt-cloud paralell provisioning
Closes#31632
* Ignore Raspbian in service.py __virtual__ (#32421)
* Ignore Raspbian in service.py __virtual__
This prevents more than one execution module from trying to load as the
service virtual module.
Refs: #32413
* pack __salt__ before loading provider overrides
We can (and should) pack here since we're just packing a reference to the
object. __salt__ needs to be available when we're loading our provider
overrides
* Fix broken __salt__ dict in provider override
Using ret.items() here sets ``__salt__`` to its items (tuple containing
function name and reference), breaking usage of ``__salt__`` inside
overridden functions.
* Merge #32293 with test fixes (#32418)
* Fix issue #11497
* Remove check for working directory presence in tests
* Fix Domainname introspection
Default value needs to be extracted from the container itself,
because dockerd set Domainname value when network_mode=host.
* Add pgjsonb_queue to queue doc index
* Pylint fixes
* Pass parser options into batch mode
Resolves#31738
* Changed the target file in file.symlink test (#32443)
* Argument name in docs should match actual arg name (#32445)
Fixes#31851
* tests.integration: bypass MacOS TMPDIR, gettempdir (#32447)
Updates 0edd532, 8f558a5.
When logging in as root over `ssh root@host`, `$TMPDIR` and
`tempfile.gettempdir()` are both set to a variation of:
```
/private/var/folders/zz/zyxvpxvq6csfxvn_n0000000000000/T/
```
When logging in as root over `sudo -i`, `$TMPDIR` is unset and
`tempfile.gettempdir()` is set to `/tmp`.
My guess is that the second case is an unintended or uncorrected omision
by Apple as they have introduced the longer, randomized temp path in a
recent version of MacOS.
* Issue #28706: Fix state user.present behavior. (#32448)
- As mentionned in issue #28706, state user.present no longer remove
user from groups if the keyword 'groups' with empty value '[]' is not
explicitly set, salt will assume current groups are still wanted.
* tests.integration: fix 4230c8a
* Move the tables of virtual modules to individual documentation pages
* Add new doc pages to toctree
* Add external ref to windows package manager docs
* Improve docstrings
* Add documentation on virtual module provider overrides to the module docs
* Clarify the scope of the provider param in states.
* Add link to provider override docs to all package providers
* Add link to provider override docs to all service providers
* Add link to provider override docs to all user providers
* dd link to provider override docs to all shadow providers
* Add link to provider override docs to all group providers
* Backport 31164 and 31364 (#32474)
* Don't send REQ while another one is waiting for response.
The message has to be removed from the queue the only *after* it's
already processed to don't confuse send() functionality that expects
empty queue means: there's no active sendings.
* Fixed zeromq ReqMessageClient destroy
* Add link to provider override docs to opkg.py
This is a companion to https://github.com/saltstack/salt/pull/32458, but
this module was not added until the 2016.3 branch, so the documentation
is being updated there for this module.
* Add documentation for some master/minion configs (#32454)
Refs #32400
Adds docs for:
- cli_summary
- event_return_queue
- event_return_whitelist
- event_return_blacklist
- file_recv_max_size
- fileserver_followsymlinks
- fileserver_ignoresymlinks
- fileserver_limit_traversal
* Automatically detect correct MySQL password column for 5.7 and fix setting passwords (#32440)
* Automatically detect MySQL password column
* Fix changing password in MySQL 5.7
* Fix lint test
* Fix unit tests (?)
They will still fail if "authentication_string" is legitimately the right column name, but I don't know what to do about that.
* Additional unit test fix
* Only unsub if we have a jid
Closes#32479
Updates 0edd532, 8f558a5.
When logging in as root over `ssh root@host`, `$TMPDIR` and
`tempfile.gettempdir()` are both set to a variation of:
```
/private/var/folders/zz/zyxvpxvq6csfxvn_n0000000000000/T/
```
When logging in as root over `sudo -i`, `$TMPDIR` is unset and
`tempfile.gettempdir()` is set to `/tmp`.
My guess is that the second case is an unintended or uncorrected omision
by Apple as they have introduced the longer, randomized temp path in a
recent version of MacOS.
