summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* CHEF-3660: Add FileUtils::Entry_ monkey patchCHEF-3660-and-windows-jenkinssdelano2012-12-042-0/+64
| | | | | This change is equivalent to the changes made to core Ruby at: - https://github.com/ruby/ruby/commit/7d89ecdc523aab7b9908501b6743da2c26a53825
* CHEF-3660: add functional tests for executing FileUtils.cp_r(:preserve) on ↵sdelano2012-12-042-1/+17
| | | | symlinks
* CHEF-3660: use FileUtils.cp_r to copy directories in deploy_providersdelano2012-12-043-5/+5
|
* Marking more tests not to run on Win2k3sersut2012-12-042-2/+2
|
* Disable link tests on Windows Server 2003 since symlink is not yet supported ↵sersut2012-12-043-1/+10
| | | | on this platform.
* Merge branch 'CHEF-3650' into 10-stableBryan McLellan2012-12-041-2/+2
|\
| * [CHEF-3650] check for whyrun using why_run key in doc formatterKishore2012-12-041-2/+2
|/
* CHEF-3555: Fix unit tests on windowsBryan McLellan2012-12-041-1/+1
| | | | | | | | I was worried about using @install_path because of its distance from this test; if someone change it to an array, it would defeat the purpose of the test. Alternately, I could have duplicated the short block that sets @install_path based on platform, but for the sake of simplicity just took the provided @install_path from the before block.
* CHEF-3555: fix knife cookbook site install when StringBryan McLellan2012-12-032-3/+13
| | | | | | | Ensure we turn cookbook_path into an Array before calling #first on it, as this does not work in Ruby 1.9+ H/T Aaron Kalin for the fix
* tests: use File.expand_path for expected key locationsdelano2012-11-291-1/+1
|
* exit from shell jenkins shell scriptBryan McLellan2012-11-291-1/+1
| | | | too much windows in that moment there; need to exit and not return
* Preserve return code from running rspecBryan McLellan2012-11-292-0/+12
|
* Add stress tests to jenkins on windowsBryan McLellan2012-11-291-1/+1
|
* Fully qualify Windows clases in stress testsBryan McLellan2012-11-291-4/+4
| | | | | CHEF-3155 moved Chef::Win32 to Chef::ReservedNames::Win32 to avoid collision with the Win32 gem namespace
* Update path for requireBryan McLellan2012-11-291-1/+1
| | | | | spec/support/prof/win32.rb became spec/support/platforms/prof/win32.rb in 1172bb6b4459d0cf0a0dc8d44ac8454a58aa4ede
* move test.xml ci output file back up to workspace directoryBryan McLellan2012-11-282-0/+2
|
* Switch template provider to FileUtils.cpBryan McLellan2012-11-271-1/+1
| | | | | | | | | This reverses the changes made in CHEF-1396. When copying a file over an existing file, the permissions of the existing file are maintained. When moving the file the permissions of the source file are preserved. Since CHEF-1396, we now set permissions after copying (formerly moving) the file, but only do so if there are permissions specified by the user to be set.
* fix binread for ruby 1.8danielsdeleo2012-11-271-0/+1
|
* remove deprecated `has_rdoc`danielsdeleo2012-11-271-1/+0
|
* fixes binmode issues on windowsdanielsdeleo2012-11-277-9/+50
|
* replace thin w/ webrick for func. test serverdanielsdeleo2012-11-277-28/+36
|
* try running functional tests firstdanielsdeleo2012-11-271-1/+1
| | | | ...should help avoid pollution of global state from unit tests
* CHEF-3411: Do not try to inspect recipes that do not existBryan McLellan2012-11-272-0/+21
| | | | | | We can't easily look at a recipe that was written in chef-shell, so we should try to open that 'file' because we will mask the real exception.
* CHEF-3568: expand path in validation_key for bootstrapBryan McLellan2012-11-202-2/+9
| | | | | | | Allows ~ in path. Some day we should probably expand paths in Chef::Config. Also expand encrypted data bag secret while we're in there.
* [CHEF-3572] Fix chefignore whitespace matchingTeemu Matilainen2012-11-203-2/+5
| | | | | | | `COMMENTS_AND_WHITESPACE` matched lines starting with word characters (\w) when it should have matched whitespaces (\s). Thus comment lines starting with spaces were not skipped but filenames with only word characters were.
* [CHEF-3597] Handle frozen strings in knife.rb chef_server_url var.Fletcher Nichol2012-11-202-1/+9
|
* Merge branch 'CHEF-3616-10-stable' into 10-stabledanielsdeleo2012-11-162-1/+26
|\
| * [CHEF-3616] add cipher field to edbi metadatadanielsdeleo2012-11-162-1/+26
|/ | | | | Adds back compat support for adding the "cipher" field to encrypted data bag v1 format.
* [CHEF-3392] fix ruby 1.8 detection of version 1 formatdanielsdeleo2012-11-151-1/+1
|
* Merge branch 'CHEF-3392-10-stable' into 10-stabledanielsdeleo2012-11-155-13/+320
|\
| * [CHEF-3392] (10-stable) compat with version 1 encrypted dbi formatdanielsdeleo2012-11-155-13/+320
|/
* Save @config[:knife] before testsBryan McLellan2012-11-131-0/+10
| | | | Stop trampling on the other tests bootstrap.
* Place script in repo for running CI testsBryan McLellan2012-11-132-0/+20
| | | | | | Chef 10 and 11 differ in directory structure so lets add a script to ease that This also exposes our testing a bit so other people know what we are doing interally.
* CHEF-3576: add raspbian as a supported platformLamont Granquist2012-11-131-0/+8
|
* CHEF-3577: Allow ssh config port and user to be usedBryan McLellan2012-11-133-5/+24
| | | | | | | | | | | Net::SSH::Multi takes a user argument which we pass, so we need to read the user from the ssh_config before passing the user and use it if applicable. We also were sending port 22 even if the user hadn't specified a port, due to a mixlib-cli default. This would override any port set in an ssh config. It isn't necessary to send a port, Net::SSH will default to 22 on its own.
* Merge branch 'immediate-converge-action-rebase-10' into 10-stabledanielsdeleo2012-11-0517-141/+46
|\
| * Rebased and merged branch immediate-converge-action onto 10-stable from master.tylercloke2012-11-0517-141/+46
| |
* | move test files from cherry-pick to correct part of treedanielsdeleo2012-11-053-0/+0
| |
* | [OC-4668] fix file rm in test teardown for ruby 1.8danielsdeleo2012-11-051-1/+1
| |
* | [OC-4668] fix constant scoping ::File vs. Filedanielsdeleo2012-11-051-2/+2
| |
* | Merge branch 'OC-4668'danielsdeleo2012-11-053-0/+499
|/
* CHEF-3574: Fix copy-pasta error in whyrun messageBryan McLellan2012-11-011-1/+1
|
* Bump version for developmentBryan McLellan2012-11-016-6/+6
|
* Release 10.16.210.16.2Bryan McLellan2012-10-266-6/+6
|
* Add thin back into the development group for WindowsBryan McLellan2012-10-251-5/+1
| | | | Eventmachine 1.0.0 installs on windows now
* [CHEF-3552] Enable whyrun support for remote_directory provider.Steven Danna2012-10-252-15/+31
| | | | | | | | | The remote_directory provider calls functions from the directory provider which is whyrun enabled. This results in the state-changing actions(namely mkdir) in Chef::Provider::Directory#action_create to actually be called *after* the actions in the Chef::Provider::RemoteDirecotry#action_create function, even if super is called at the beginning of the latter function.
* Don't get checksum for create_if_missingBryan McLellan2012-10-251-1/+3
| | | | | We don't need the checksum in create for create_if_missing because we're going to create because we already know the file doesn't exist.
* [CHEF-3561] add template context for template errorsdanielsdeleo2012-10-252-1/+37
| | | | | ResourceFailureInspector will now notice TemplateErrors and use them to display the relevant lines from a template.
* CHEF-3554: limit unix only specsBryan McLellan2012-10-253-132/+146
| | | | | Because we're hacking out functionality to fix CHEF-3554, we need to skip the related tests as well. These must be fixed for CHEF-3557.
* added functional test for cookbook_file providertimh2012-10-251-0/+51
|
View Lorry Controller Status