| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
| |
According to `man shadow` on Solaris 10 and 11, an account is only
locked if the first four characters in the password field are
`*LK*`. The same is true for SmartOS (the OS for which the initial code
was contributed).
Signed-off-by: Tom Duffield <tom@chef.io>
|
|
|
|
|
|
|
|
|
| |
On Solaris, the 'shadow' database does not exist within `getent`, so the
checking for the username there won't return accurate results. As the
Solaris provider assumes user management via /etc/shadow, we can very
easily parse the contents of the file directly.
Signed-off-by: Tom Duffield <tom@chef.io>
|
| |
|
|\
| |
| | |
Core: set use_inline_resources in package superclass
|
| |
| |
| |
| |
| |
| |
| | |
we have too much collaboration and state in our objects to write
pure unit tests.
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| |
| |
| |
| |
| |
| |
| | |
allows subclasses to use_inline_resources (hab_package
will most likely want this)
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| | |
|
|\ \
| |/
|/| |
Use `all?` instead of `any?` to check for SUSE users
|
| |
| |
| |
| |
| |
| |
| | |
Because `Provider::Group::Suse` expects users to exist, we need to mock
out the `Etc` call that checks for our non-existent spec users.
Signed-off-by: Tom Duffield <tom@chef.io>
|
|/
|
|
|
|
| |
Using `all?` will return `false` when the array is empty.
Signed-off-by: Tom Duffield <tom@chef.io>
|
| |
|
|\
| |
| | |
Fix copyrights and update a few more Opscode references to Chef
|
| |
| |
| |
| | |
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| |
| |
| |
| |
| |
| | |
I think this happened when we merged the Opscode and Chef copyrights
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\ \
| | |
| | | |
Fix a typo in the apt_repository spec
|
| | |
| | |
| | |
| | | |
Signed-off-by: Tim Smith <tsmith@chef.io>
|
|\ \ \
| | | |
| | | | |
Use property vs. attribute in yum_repository
|
| |/ /
| | |
| | |
| | |
| | |
| | | |
We just auto replace the attribute call with a property call here. We might as well use the right one to start with.
Signed-off-by: Tim Smith <tsmith@chef.io>
|
| | | |
|
|\ \ \
| | | |
| | | | |
Address test failures in SUSE 11.4 and 12.1
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The expected behavior of chef, as defined by our tests, is that if a
user specified in a group resource does not exist than chef should raise
an error. On suse, we use groupmod which will not raise an error if you
attempt to add a non-existent user to a group. This change adds in a
simple check that will raise an error if the specified does not exist.
Signed-off-by: Tom Duffield <tom@chef.io>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The `usermod` binary in SUSE 12.1 no longer throws an error if you
attempt to unlock a user account that has no password.
Signed-off-by: Tom Duffield <tom@chef.io>
|
| | | | |
|
|\ \ \ \
| | | | |
| | | | | |
fixes for rubocop 0.44.1 engine
|
| | |_|/
| |/| |
| | | |
| | | |
| | | |
| | | | |
a few cops we have enabled got a little more accurate/stricter
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| | | | |
|
|\ \ \ \
| |_|/ /
|/| | | |
Attributes: add tests for regression in 12.0 fixed recently
|
|/ / /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Tests for the issue that led up to
https://github.com/chef/chef/issues/5447
The Hash tests actually pass against 12.13.37 so it appearly that
only broke the Immutable array objects in 12.0. This both explains
why we didn't have any existing tests that caught it, and why it went
uncaught for so long. It also adds some ammunition to the argument
that fixing this in 12.14.x was not a major breaking change to users
that had been relying on the buggy 12.0 behavior.
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| | | |
|
|\ \ \
| |/ /
|/| | |
convert svn_password to property + make sensitive
|
| | |
| | |
| | |
| | |
| | |
| | | |
should property-ize the rest of this as well
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| | |
| | |
| | |
| | | |
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| | | |
|
|\ \ \
| |/ /
|/| | |
Cron Provider: Eliminate crontab rewrite that happens with every chef-client run
|
| | |
| | |
| | |
| | | |
Signed-off-by: Peter Walz <pnw@umn.edu>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Fixes a side effect of #5127 where if a `cron` resource sets MAILTO, PATH,
SHELL, or HOME, it gets updated with every chef-client run.
Signed-off-by: Peter Walz <pnw@umn.edu>
|
| | | |
|
|\ \ \
| | | |
| | | | |
Update EC2 tags for internal standards compliance.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
[ci skip]
Signed-off-by: Ryan Hass <rhass@chef.io>
|
| | | | |
|
|\ \ \ \
| | | | |
| | | | | |
[Knife] Fix deprecation warning when creating data bag items
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Fixes chef/chef-vault#215
Signed-off-by: Thom May <thom@may.lt>
|
| | | | | |
|
|\ \ \ \ \
| | | | | |
| | | | | | |
remove dependency on Chef.run_context global state
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| | |/ / /
| |/| | |
| | | | |
| | | | |
| | | | |
| | | | | |
introduce run_context.root_run_context that avoids global state
Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
|
| | | | | |
|
|\ \ \ \ \
| | | | | |
| | | | | | |
Bump omnibus deps once more.
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Signed-off-by: Ryan Hass <rhass@chef.io>
|