summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Fixing #1631 - adding /sbin and /usr/sbin to PATHLuke Kanies2009-03-242-0/+18
| | | | | | | This is a trivial fix but seems to crop up more often than it should. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixed #2004 - ssh_authorized_key fails if no target is definedFrancois Deppierraz2009-03-244-44/+48
| | | | | | | This commit depends on 7f291afdacf59f762c3b78481f5420ec8919e46d (fixing #1629) which was cherry-picked from master. Signed-off-by: Francois Deppierraz <francois@ctrlaltdel.ch>
* Changelog entries for #1629 and #2004Francois Deppierraz2009-03-241-0/+4
| | | | Signed-off-by: Francois Deppierraz <francois@ctrlaltdel.ch>
* Fix Bug #1629Francois Deppierraz2009-03-242-12/+116
| | | | | A refactoring of ssh_authorized_key parsed provider was needed and tests were improved. flush method has been split for clarity.
* Fix #1807 - make Puppet::Util::Package.versioncmp a module functionBrice Figureau2009-03-242-0/+23
| | | | Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fix #1829 - Add puppet function versioncmp to compare versionsBrice Figureau2009-03-242-0/+39
| | | | Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fixed install.rb typoJames Turnbull2009-03-241-1/+1
|
* Updated lib install permissions to 0644James Turnbull2009-03-241-1/+1
|
* Fixes incorrect detail variable in OS X version check, re-patches ralsh to ↵Nigel Kersten2009-03-203-3/+9
| | | | work with Facter values and adds error check for missing password hash files.
* Fix #1828 - Scope.number? wasn't strict enough and could produce wrong resultsBrice Figureau2009-03-142-3/+17
| | | | | | | Some invalid numbers were treated as numbers and conversion to Integer was failing returning 0 (for instance 0.24.7). Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fixed #2000 - No default specified for checksumJames Turnbull2009-03-092-1/+9
|
* Fixing change printing when list properties are absentLuke Kanies2009-03-072-1/+9
| | | | | | | They were throwing an exception when the 'is' value was 'absent'. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixed #2026 - Red Hat ignoring stop methodJames Turnbull2009-03-072-4/+6
|
* Bring in the documentation changes from the master branchBryan Kearney2009-03-041-3/+2
|
* Added a force option to ensure the change is always applied, and call augeas ↵Bryan Kearney2009-03-043-146/+197
| | | | twice to reduce the chance that data is lost
* Backport the fix for #1835Bryan Kearney2009-03-042-4/+102
|
* First cut at the not running if augeas does not change any of the ↵Bryan Kearney2009-03-042-34/+92
| | | | underlieing files
* Bug 1948: Added patch by jab to support the correct ins syntax. Updated the ↵Bryan Kearney2009-03-041-5/+3
| | | | test cases as well
* Fixing #1991 - ldap booleans get converted to booleansLuke Kanies2009-03-052-0/+45
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Refactored a method: extracted about five other methodsLuke Kanies2009-03-051-39/+50
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* dbfix - fix typo and close another possible inconsistencyBrice Figureau2009-03-041-3/+37
| | | | | | | | | | | This patch fixes a typo in one request producing a SQL error. It also close an issue where param_values could point to inexistant param_names after the deduplications of those. It does this by rewriting the param_values to point to the highest id of the param_names of the same name. The same operation is performed on the tags. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fix #2010 - add protection code for some storeconfig corruptionBrice Figureau2009-03-041-14/+28
| | | | | | | | | | | | | | This patch adds a more robust and self-healing storedconfig in case of logically corrupted database as the one in #2010 (where multiple resources of same references are present in the database for the same host). The problem is that the resources are stored in a hash with the resource ref as the key, so we collapse resource of different id but same reference. The patch fixed this by using a hash by resource id, and maintaining a list of old extraneous resource in the db that are removved after the storeconfig pass. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Further fix to #1910James Turnbull2009-03-041-1/+1
|
* Fixing #2013 - prefetching had a mismatch between type and titleLuke Kanies2009-03-042-1/+16
| | | | | | | | | The ParsedFile types seem to be the main one that suffers from this, but the transactions were using the resource titles, not names, so resources were often not getting prefetched correctly. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixed to rake tests for reductivelabs buildJames Turnbull2009-03-041-3/+3
|
* Fixed report reference pageJames Turnbull2009-03-031-0/+1
|
* Fixing broken 0.24.x tests in test/.Luke Kanies2009-02-283-4/+10
| | | | | | | | | | | | | These tests once again largely were caused by /usr/sbin not being in the path and by ~ not being writable. The only tests still failing are Rails tests, and my guess is that they're all failing because of the recent work by Brice. They should probably just be removed. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixing every failing test I can find on the build server.Luke Kanies2009-02-2814-7/+42
| | | | | | | | | | | | | | | | All but one of these tests is fixed by: * Stubbing Puppet.settings.use * Adding /usr/sbin to PATH The only other one was the package integration test, which stupidly assumed a default was specified in the test. The fix here is twofold: Remove that assumption (the test is now 'pending' if no default is available), and add a default for Ubuntu. (The default is in the test, not the default package provider - that is, it's testing the default package provider, thus it can't rely on that information.) Signed-off-by: Luke Kanies <luke@madstop.com>
* This script fixes the most common issues with inconsistentJames Turnbull2009-02-282-0/+100
| | | | | | | | | | | | | | storeconfigs database (including duplicate resources record, duplicate param_values records, dangling records...). Usage: stop all puppetmasters backup your database! % cat ext/dbfix.sql | mysql puppet relaunch all puppetmasters Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com> ---
* Make puppetd --waitforcert option behave as documented:Paul Lathrop2009-02-282-1/+19
| | | | | | | | "You can turn off waiting for certificates by specifying a time of 0." Also add a test to ensure we catch any future regression of this behavior. Signed-off-by: Paul Lathrop <paul@tertiusfamily.net>
* Adding a performance optimization to the FileCollection.Luke Kanies2009-02-281-3/+5
| | | | | | | I saw about a 7x speed increase when adding this simple hash to speed up the file index lookup. Signed-off-by: Luke Kanies <luke@madstop.com>
* Using the FileCollection where appropriate.Luke Kanies2009-02-2814-12/+60
| | | | | | | | | | | | | This commit just replaces the :file and :line accessors with the use of the new FileCollection Lookup module. This should mean that we've normalized all file names in a given process, which *might* have drastic RAM improvements. For initial simplicity, I've gone with a single global collection of file names, but it's built so it's easy to use individual file collections instead. Signed-off-by: Luke Kanies <luke@madstop.com>
* Adding a FileCollection and a lookup module for it.Luke Kanies2009-02-284-0/+122
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixed #1963 - Failing to read /proc/mounts for selinux kills file downloadsJames Turnbull2009-02-282-1/+5
|
* Fixed #2025 - gentoo service provider handle only default init levelJames Turnbull2009-02-272-1/+3
|
* Fixed #1910 - updated logcheckJames Turnbull2009-02-262-1/+4
|
* Updated useradd.rb managehome confine to include other RH-like distributionsJames Turnbull2009-02-251-1/+1
|
* Use Puppet.debug instead of own debug flagPeter Meier2009-02-221-7/+5
| | | | | | It's better to use puppet's logging environment than an own. Especially if the default is quite verbose and can't be set by config flag.
* Fixing #558 - File checksums no longer refer to 'nosum'Luke Kanies2009-02-201-34/+0
| | | | | | | It was only used in a method that apparently was not in use any more. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixing #1871 once and for all - contents are never printedLuke Kanies2009-02-183-8/+43
| | | | | | | | | They were still being printed in noop mode. The fix was to use is_to_s and should_to_s methods, rather than 'change_to_s'. Signed-off-by: Luke Kanies <luke@madstop.com>
* Minor fix to launchd testsJames Turnbull2009-02-151-2/+2
|
* Fix #1972 - ActiveRecord fixes resulted in broken testsBrice Figureau2009-02-151-3/+4
| | | | | | | Fix #1930 was not complete, with some typos and misuse of unused code paths in the regular puppetmaster use case. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fix spec test for launchd service provider to work with new service status ↵Nigel Kersten2009-02-151-1/+9
| | | | | | method and add two new status tests. Now ensures that given the correct input, the correct status for the launchd service is returned.
* Fixing a test I broke in commit:"897539e857b0da9145f15648b6aa2ef124ec1a19".Luke Kanies2009-02-141-3/+0
| | | | | | I hadn't removed the call to 'prefetch' on the instance in the test. Signed-off-by: Luke Kanies <luke@madstop.com>
* Removing a no-longer-valid test.Luke Kanies2009-02-141-40/+0
| | | | | | | The fix for #1884 removed this no-longer-needed feature, so this test is now unnecessary. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixing password validation to support symbols.Luke Kanies2009-02-142-5/+9
| | | | | | | My fix for #1920 broke when you set :absent or :present. Signed-off-by: Luke Kanies <luke@madstop.com>
* Only backing up within parsedfile when managing filesLuke Kanies2009-02-143-8/+19
| | | | | | | | | | | I was getting failing tests because I was using non-files for testing and they didn't back up the same, not surprisingly. This moves the 'backup' method to the :flat filetype and then only backs up if the filetype supports it. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixing a syntax error in the up2date providerLuke Kanies2009-02-141-1/+1
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixing a test broken by a log demotionLuke Kanies2009-02-141-2/+2
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Cleaned up variable names to be more sane, clarified error messages and ↵James Turnbull2009-02-142-6/+9
| | | | fixed incorrect use of 'value' variable rather than 'member'.