Permalink
Switch branches/tags
Commits on Jun 18, 2014
  1. * README.md: add pointer to JIRA

    lutter committed Jun 18, 2014
Commits on May 28, 2014
  1. Remove old code and replace it with updated README.md

    lutter committed May 28, 2014
    QUite a few people were getting confused between this repo and the new and
    maintained https://github.com/puppetlabs/razor-server; since there is some
    interest in the historical artifacts of the repo, we keep it public but in
    a way that should minimize confusion.
Commits on Aug 21, 2013
Commits on Aug 19, 2013
  1. Merge pull request #505 from toabctl/fix-gemspec-homepage

    slippycheeze committed Aug 19, 2013
    Use correct homepage in gemspec
Commits on Aug 18, 2013
  1. Use correct homepage in gemspec

    toabctl committed Aug 18, 2013
Commits on Jul 15, 2013
  1. Import "The road forward for Razor" as the primary readme

    slippycheeze committed Jul 15, 2013
    This brings information about what is going on with Razor into the forefront
    of visibility.  Old information is still preserved in a separate file.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>
Commits on Jun 26, 2013
  1. Merge pull request #492 from MosesMendoza/packaging/master/update_pac…

    haus committed Jun 26, 2013
    …kaging_loader
    
    (packaging) Use the packaging loader for tasks
  2. Merge pull request #491 from MosesMendoza/packaging/master/remove_one…

    haus committed Jun 26, 2013
    …iric
    
    (packaging) Remove Ubuntu Oneiric from build targets
Commits on Jun 25, 2013
  1. (packaging) Remove Ubuntu Oneiric from build targets

    MosesMendoza committed Jun 25, 2013
    Oneiric is EOL, and we are no longer building for it. This commit removes the
    Oneiric target from the build list for Razor.
    
    Signed-off-by: Moses Mendoza <moses@puppetlabs.com>
  2. (packaging) Use the packaging loader for tasks

    MosesMendoza committed Jun 24, 2013
    The packaging repo now uses an explicit loader which handles loading the
    various rake tasks in packaging. This commit updates the Razor Rakefile to
    use the loader instead of a glob of the ext/packaging/tasks directory. We move
    this load into the rescue LoadError block because the packaging repo won't
    always be there, but if it is, we want to load it.
    
    Signed-off-by: Moses Mendoza <moses@puppetlabs.com>
Commits on Jun 17, 2013
  1. Merge pull request #484 from mhanco/bug/master/slice-log-class-file-t…

    slippycheeze committed Jun 17, 2013
    …ail-method-size-undefined
    
    Fixed size undefined error in tail method extension of class File.
Commits on Jun 5, 2013
  1. Merge pull request #479 from cr3/478

    slippycheeze committed Jun 5, 2013
    (#478) /etc/rc.local should remain a symlink on Red Hat
Commits on May 29, 2013
  1. Merge pull request #482 from daniel-pittman/bug/master/adapt-to-chang…

    slippycheeze committed May 29, 2013
    …ed-internal-machine-name-in-acceptance-tests
    
    Update acceptance tests after internal machines were renamed.
  2. Update acceptance tests after internal machines were renamed.

    slippycheeze committed May 29, 2013
    This updates the acceptance test performance suite to reflect internal machine
    name changes on the Puppet Labs network.  This restores the acceptance tests
    to their fully functional, original state.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>
Commits on May 28, 2013
  1. Merge pull request #481 from daniel-pittman/bug/master/adapt-to-new-p…

    slippycheeze committed May 28, 2013
    …uppet-repo-layout-for-acceptance-tests
    
    Update CI testing Puppet installation script repo setup
  2. Update CI testing Puppet installation script repo setup

    slippycheeze committed May 28, 2013
    Puppet repositories have recently changed, resulting in a need to update the
    scripts we use in the CI system to reflect that.  This allows Puppet to be
    installed as part of the test regime.
    
    This has not changed the broker installation process in any way, since that is
    not using any of the repositories that have changed structure.
    
    There are not functional changes to Razor as a result of this work, as it is
    strictly related to Puppet Labs internal CI infrastructure.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>
Commits on May 13, 2013
  1. (#478) /etc/rc.local should remain a symlink on Red Hat

    Marc Tardif committed May 13, 2013
    This change adds the --follow-symlinks option to the sed command in the
    os_complete.erb template of the redhat model. This is necessary because
    the file /etc/rc.local on Red Hat should be a symlink to rc.d/rc.local
    but, without this option, it becomes a regular file.
    
    Note that the same change should not be necessary in Ubuntu and Debian
    models because /etc/rc.local is not a symlink on those operating systems.
    
    Closes #478
Commits on May 10, 2013
  1. Merge pull request #475 from daniel-pittman/bug/master/474-revert-add…

    tjmcs committed May 10, 2013
    …itional-partitioning-support
    
    Revert "add posibility to use 'autopart' or 'part' in redhat's kickstart...
  2. Revert "add posibility to use 'autopart' or 'part' in redhat's kickst…

    slippycheeze committed May 10, 2013
    …art"
    
    This reverts commit 104b066.  Making it possible for users to customize the
    partition layout is highly undesirable with the other committers, to
    my surprise.
    
    Given that, this is being reverted.  Users who want this capability are
    expected to build their own model to implement it.
Commits on May 9, 2013
  1. Merge pull request #474 from faja/feature/add_part_in_redhat_model

    slippycheeze committed May 9, 2013
    add posibility to use 'autopart' or 'part' in redhat's kickstart
Commits on May 7, 2013
  1. Merge pull request #472 from cr3/471

    slippycheeze committed May 7, 2013
    (#471) Inconsistent assumption about @node_ip in models
  2. (#471) Inconsistent assumption about @node_ip in models

    Marc Tardif committed May 7, 2013
    This change fixes the broker_agent_handoff method in the models that
    potentially used an empty ipaddress option when calling the agent_hand_off
    method in the broker. Now, the agent_hand_off method is only called when
    the @node_ip is not empty and still reports a useful error message when
    it is empty.
    
    Closes #471
Commits on May 3, 2013
  1. Merge pull request #466 from cr3/464

    slippycheeze committed May 3, 2013
    (#464) Facter should only be installed if necessary in os_boot.erb
Commits on Apr 25, 2013
  1. (#464) Facter should only be installed if necessary in os_boot.erb

    Marc Tardif committed Apr 25, 2013
    This change adds the option --enablerepo=base to the yum install call
    in os_boot.erb to make sure the base repo is always enabled. This is
    necessary because the factor package from the puppetlabs repo depends
    on other packages like virt-what for example which are only available
    in the base repo.
    
    The change should not introduce any regressions because it simply
    makes sure that the base repo is enabled. If it is enabled already,
    this will not make any difference.
    
    This pull request fixes issue #464.
Commits on Apr 19, 2013
  1. Remove Ubuntu Raring package builds temporarily

    slippycheeze committed Apr 19, 2013
    Ubuntu Raring builds are currently not on our core "supported" list
    of platforms; this removes the efforts to build packages for them.
    
    It will return, eventually, but not without some additional planning.
Commits on Apr 18, 2013
  1. Merge pull request #460 from daniel-pittman/feature/master/459-bring-…

    slippycheeze committed Apr 18, 2013
    …option-defitition-into-code-rather-than-yaml-files
    
    
    #459: bring option defitition into code rather than yaml files
  2. Merge pull request #458 from daniel-pittman/feature/master/326-killal…

    slippycheeze committed Apr 18, 2013
    …l-node-is-not-awesome-as-shutdown-strategy
    
    Use `get_node_instance_info` to target stopping node processes
Commits on Apr 16, 2013
  1. Fix mismatched option `invert` with code `inverse`

    slippycheeze committed Apr 16, 2013
    An earlier fix caught most, but not all, of the mismatching uses of `invert`
    and `inverse` through the code.  Jan Meyer <jan.meyer@m-bs.de> correctly noted
    one of the places that was missed was the external option data files.
    
    This, in part, is the motivation for pulling that data into the Ruby code -
    having established that it is harder to discover that in a separate file, when
    you search for a string, I wanted to tackle the root cause.
    
    Ultimately, though, I need to apply the fix that Jan supplied - and having
    rewritten the code this applies the concept rather than the delta he offered.
    
    This supplants #451, which made the change in the YAML files.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>
  2. Turn YAML option data into Ruby code option data

    slippycheeze committed Apr 16, 2013
    This imports the YAML option definitions into Ruby code, avoiding the need to
    parse and decode an external file before we can print option data.  This also
    makes it harder - though, obviously, still entirely possible - to miss the
    existence of that data when updating the code.
    
    There is no behavioural change in the product from this.  The same data, in
    the same form and order, is supplied to the rest of the system.  Only the
    source is modified.
    
    This closes #459, as the data now lives in Ruby.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>
  3. Use `get_node_instance_info` to target stopping node processes

    slippycheeze committed Apr 16, 2013
    Previously, the Razor daemon tried to shut down the API instances using
    `killall` on anything named `node` - which failed on two counts, being that it
    would kill too many things, and that it didn't work on some distributions
    which ended up naming the process `nodejs` instead.
    
    This replaces that with a more targeted process that kills only the nodejs
    processes directly associated with Razor.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>
  4. Rename `load_option_items` to `command_option_data`

    slippycheeze committed Apr 16, 2013
    This renames the `load_option_items` method to `command_option_data`, so that
    it is named after what you get from invoking it, rather than what it did
    internally to retrieve that information.
    
    This commit has no other functional change.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>
  5. Inline discovery of the YAML file containing option items

    slippycheeze committed Apr 16, 2013
    The discovery of the YAML file containing option items was only used through
    the `load_option_items` method on the base slice class.
    
    Since this was not really API in any meaningful sense, this commit inlines the
    behaviour and reduces our footprint.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>
  6. Eliminate a dead call to `print_command_help`

    slippycheeze committed Apr 16, 2013
    The BMC code had a commented out invocation of `print_command_help`; this
    eliminates it, to make clearer the behaviour of the system.
    
    Signed-off-by: Daniel Pittman <daniel@rimspace.net>