| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
Convert this to a true property. It required updating the specs a tiny bit since key gets returned as a state value now (as it should).
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\
| |
| | |
prune remotes with git fetch, fixes #3929
|
| |
| |
| |
| | |
Signed-off-by: Roland Moriz <roland@moriz.de>
|
|\ \
| | |
| | |
| | |
| | | |
RoboticCheese/roboticcheese/package-locking-nomethoderror
Fix NoMethodError when (un)locking single packages in apt and zypper
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
These resources don't coerce `package_name` to an array so would raise an
exception when trying to call `name.all?` when `package_name`, and thus `name`,
was a string.
Signed-off-by: Jonathan Hartman <j@hartman.io>
|
|\ \ \
| | | |
| | | | |
[Debian/Ubuntu] Detect init script service levels
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Debian removed the no-op flag from update-rc.d, and so we're now
unlinking services rather than just querying them. Just parsing the
scripts ourselves is much safer.
Fixes: #7119
Signed-off-by: Thom May <thom@chef.io>
|
|/ / /
| | |
| | |
| | | |
Signed-off-by: Stuart Preston <stuart@chef.io>
|
|/ /
| |
| |
| | |
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
|/
|
|
| |
Signed-off-by: Erik Gomez <e@eriknicolasgomez.com>
|
|\
| |
| | |
[CHEF-7026] Rewrite portage package provider candidate_version determination and fix tests
|
| |
| |
| |
| | |
Signed-off-by: Gordon Malm <gmalm+github-chef-chef-2018@engineyard.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Reference:
https://github.com/chef/chef/issues/7026
03fffdcd91b71ee0ff346310c5c9616976ca3f1b
Signed-off-by: Gordon Malm <gmalm+github-chef-chef-2018@engineyard.com>
|
|\ \
| | |
| | | |
Avoid lookups for rights of 'LocalSystem' in windows service
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
LocalSystem is a special account for the service subsystem, and the security
subsystem doesn't know about it. It inherits rights from BUILTIN\Administrators
so we don't need to check it for SeServiceLogonRight. Even if we look up System
it wouldn't show up as it gets that right from hidden membership in
BUILTIN\Administrators.
Signed-off-by: Bryan McLellan <btm@loftninjas.org>
|
|\ \ \
| |/ /
|/| | |
windows_feature_dism: Be case insensitive with feature names
|
| | |
| | |
| | |
| | |
| | |
| | | |
Dism here does not support case insensitive feature installs. This isn't a user behavior change sine they always needed the feature to have the right case here.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
We took what they user gave us and compared it to the list of available
packages from dism.exe. In doing so we were case sensitive, but the CLI
doesn't actually care. This just downcases what the user gives us and
what dism gives us so it doesn't matter.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\ \ \
| |/ /
|/| | |
Setting nil to properties with implicit nil sets default value
|
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
- This makes converting core resources to properties safer
- This makes it easier to apply wrapping properties to subresources
property :foo, String, default: "foo"
This is where the change lies, and writing a nil here will now actually
write a "foo" to the variable.
property :foo, [ String, nil ], default: "foo"
This is unchanged. Writing nil writes nil.
property :foo, String
Technically this is changed, since it writes the default value, but
since nil.equal?(nil) in a very deep way no behavior changes.
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This fixes how we parse out the arrays and adds testing to make sure it's doing what we want.
Plus it properly continues on when the user has removed all local feature but specified an external source of those via the registry. That was a reported issue that came in on the cookbook after the rewrite.
Lastly this also fixes a bad method call to the dism method that came over when diffing from the cookbook.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\ \
| | |
| | | |
RFC-102: Deprecation warning in resources
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* `deprecated_property_alias` allows the resource author to provide
transition from old properties to new ones with a deprecation
warning.
* The `deprecated` option on a property emits a deprecation warning.
* The `deprecated` method on a resource takes a message, but does not
yet emit a deprecation warning.
Signed-off-by: Thom May <thom@chef.io>
|
| | |
| | |
| | |
| | | |
Signed-off-by: Thom May <thom@chef.io>
|
| | |
| | |
| | |
| | | |
Signed-off-by: Thom May <thom@chef.io>
|
|\ \ \
| | | |
| | | | |
Remove path helpers from window_pagefile
|
| |/ /
| | |
| | |
| | |
| | |
| | | |
I backported locate_sysnative_cmd initially, but after chatting with Stuart this isn't something we need anymore. This removes it and instead uses wmic.exe directly, which in testing works just fine. This also removes the windows friendly path helper which is really just a gsub on the slashes. Instead I changed the coerce to get us what we need in terms of path formats. I copied this to a Windows 2016 box and it works fine with both formats of paths. I'm making this same change to the windows cookbook.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|/ /
| |
| |
| |
| |
| | |
Copied as-is from the cookbook
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |
| |
| |
| |
| |
| | |
This allows us to have windows_ad_* in the future
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |
| |
| |
| | |
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |
| |
| |
| |
| |
| | |
Rename the property to reboot to align with Chef naming. Allow delayed, immediate, or never for reboot actions. Using the actions from reboot was pretty confusing so I used notification names which seems to lineup nicely.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |
| |
| |
| |
| |
| | |
Copied with permission from John Snow's win_ad cookbook. It's been reworked to not mess with dns servers and use powershell_out + sensitive. Some of those changes are getting pushed upstream, but the name is different and there's no expectation of compatibility here.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\ \
| | |
| | | |
Add sysctl_param resource from the sysctl cookbook
|
| | |
| | |
| | |
| | | |
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| | |
| | |
| | |
| | |
| | |
| | | |
Copied from the cookbook with modifications made there
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| | |
| | |
| | |
| | |
| | |
| | | |
I guess they use .d directory structures in sudo, but that's about it.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |/
|/|
| |
| |
| |
| | |
Simple rename
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\ \
| | |
| | | |
rubocop fixes from engine bump to 0.54.0
|
| |/
| |
| |
| |
| |
| | |
some cops got a little stricter.
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
|\ \
| |/
|/| |
Save the node's UUID as an attribute
|
| |
| |
| |
| |
| |
| |
| | |
We generate the UUID as part of the data collector report, but we didn't
make that available to the node or the chef server otherwise.
Signed-off-by: Thom May <thom@chef.io>
|
| |
| |
| |
| |
| |
| | |
This converts the Chef resources to shell_out and adds new properties for dealing with edge case installations. It also adds validation messaging / testing and lazy evals the stuff that uses the mixin.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |
| |
| |
| |
| |
| | |
Pull in resources from the dmg and homebrew cookbooks
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |
| |
| |
| |
| |
| | |
We set a lot of nil defaults and did coersions here that are no longer necessary. Since we did accept nil in a lot of these properties at one point we need to keep it so I wrote a note to future me about that. I also added some specs for the remaining coersion and the defaults.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\ \
| | |
| | | |
Fail with a warning if users specify apt/yum/zypper repos with slashes
|
| | |
| | |
| | |
| | |
| | |
| | | |
We can't write out a filename like foo/bar.repo so we should properly warn the user if they try to give us that. There's probably other things, but this one makes particular sense for apt since people want to use the slash in PPA repo names.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|/ /
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This resolves issue #6836 which explains the issue in great detail.
TLDR: Packages that are installed as deps on another package have a
different string when you get info on them now. This updates the regex
to detect both. Based on the zypper changelog I believe this was
introduced in 1.13.17.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |
| |
| |
| |
| |
| | |
meaning that we can have inheritable metadata
Signed-off-by: Thom May <thom@chef.io>
|
|/
|
|
|
|
| |
We actually set the defaults for these in the subversion property, but the actual properties are in scm. Move them to subversion since they are subversion specific.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\
| |
| | |
Add the sudo resource from the sudo resource
|