Re: [Puppet-dev] Class ProviderAixuseradd is already defined in Puppet::Type::User

On Tue, Nov 30, 2010 at 8:33 PM, Luke Kanies wrote: >> I do not require directly any puppet module. > > You're directly requiring the aixobject parent class, which shouldn't be > necessary.  Can you try removing that and see if it fixes it? It started to work again without change, but I will re

[Puppet-dev] Re: Bug #5423: This moves the home directory property before the uid property, thus minimizing room for damage when usermod is in use.

Jonathan wrote: > James- > > I literally said those exact words. But on a whim I changed then order > and it works correctly, every time. I haven't had a chance to dig deeper > yet, so I'm not sure if the issue is in the namespace provider, but the > patch definitely fixes it. > > Try out the pat

[Puppet-dev] Re: Bug #5423: This moves the home directory property before the uid property, thus minimizing room for damage when usermod is in use.

Jonathan wrote: > James- > > I initially thought the same thing. Here's some steps to reproduce the > issue: > > 1. Make a simple manifest, similar to the one below: > > user { "testy": > ensure => present, > uid => "2345", > gid => "2345", > groups => ["testy"],

Re: [Puppet-dev] [PATCH/puppet 1/1] Bug #5423: This moves the home directory property before the uid property, thus minimizing room for damage when usermod is in use.

James Turnbull wrote: > From: Jonathan Boyett > Jonathan I don't see why the order of the properties would make any difference to the way this applies. They are not executed in sequence from what I understand. Regards James Turnbull -- Puppet Labs - http://www.puppetlabs.com C: 503-734-8571

Re: [Puppet-dev] Best way to parse /etc/services

On Nov 23, 2010, at 10:35 AM, Stefan Schulte wrote: > On Tue, Nov 23, 2010 at 12:36:07AM -0600, Luke Kanies wrote: >> If you invert it, it works better: >> >> port { '22/udp': label => 'telnet' } >> > > Havent thought of that an it looks pretty good for me. The duplication > doesnt really bothe

Re: [Puppet-dev] [PATCH/puppet 4/4] (#5274) Using Propery::OrderedList for host_alias

Stefan, This 4th commit didn't get in with your previous change set since it was added after we initially looked at the code. Also, it really doesn't have anything to do with ticket #5274 which is about inline comments, so this commit should really be part of another ticket. We've opened a new tic

Re: [Puppet-dev] [PATCH/puppet 1/1] (#2495) Better value validation for sshkey

+1 On Sun, Nov 21, 2010 at 12:02 PM, Stefan Schulte wrote: > As mentioned in the ticket it is not obvious that aliases do not belong > in the resourcename but have to be specified with the property > "host_aliases". On the puppet-user list I saw someone using this as a > resource > >  @@sshkey {"

Re: [Puppet-dev] [PATCH/puppet 1/2] Maint: Refactor tests to use .indirection.

On Nov 30, 2010, at 10:39 AM, Paul Berry wrote: > On Mon, Nov 29, 2010 at 10:47 PM, Luke Kanies wrote: > What's the reason for this change? > > I assume it's related to the default route stuff, but I'd like to understand > a bit more about this direction. In particular, I thought the default r

Re: [Puppet-dev] Best way to parse /etc/services

If you wanted to have the service as the resource title and pass the protocols as an array like this: port { 'telnet', protocols => ['tcp', 'udp'], number=> '22', port_aliases => ['alias1','alias2'], } you can use a template that iterates on `protocols.each do |p|` and returns multiple

Re: [Puppet-dev] [PATCH/puppet 0/8] Experimental Plugable Puppet Authentication System

Brice, Thanks for your contribution. Matt Robinson and I looked your patch over, and have a few comments. I like the idea, and the code is reasonable, but we'd like to see another example auth plugin, using this new system. Without that, it's hard to design a suitable authentication API. Addition

Re: [Puppet-dev] [PATCH/puppet 1/1] (#5422) Ship auth.conf as part of installing from source Add --configdir which defaults to /etc/puppet Add --no-configs to suppress config file installation. We're

On Tue, Nov 30, 2010 at 3:02 PM, Jacob Helwig wrote: > Forget a linebreak in the subject there somewhere? Somewhere :) > > Looks like there's a few places where it's trying _extra_ hard to ensure > directories where it really only need to try, but that's more a > statement about the install.rb i

Re: [Puppet-dev] [PATCH/puppet 1/1] (#5422) Ship auth.conf as part of installing from source Add --configdir which defaults to /etc/puppet Add --no-configs to suppress config file installation. We're

Forget a linebreak in the subject there somewhere? Looks like there's a few places where it's trying _extra_ hard to ensure directories where it really only need to try, but that's more a statement about the install.rb in general. It might not hurt to come through here with a broom at some point

Re: [Puppet-dev] Re: (#4974) Allow tags=main to apply all resources.

> Also, what do you mean about cross-stage dependencies being broken? Sorry, misinterpreted something Nan said. It works like it should (ie, creating circular dependencies when a resource in a later stage requires a resource in an earlier stage). -Hunter -- You received this message because you

[Puppet-dev] [PATCH/puppet 4/5] Maint: Modified tests of indirector.save to call the indirection directly.

This change replaces calls to .save with calls to .indirection.save(). This makes the use of the indirector explicit rather than implicit so that it will be easier to search for all indirector call sites using grep. This is an intermediate refactor on the way towards allowing indirector calls to

[Puppet-dev] [PATCH/puppet 5/5] Maint: Modified uses of indirector.save to call the indirection directly.

This change replaces calls to .save with calls to .indirection.save(). This makes the use of the indirector explicit rather than implicit so that it will be easier to search for all indirector call sites using grep. This is an intermediate refactor on the way towards allowing indirector calls to

[Puppet-dev] [PATCH/puppet 1/5] Maint: Swap the order of arguments to Indirection#save

The first argument was often nil, and the second was mandatory. Signed-off-by: Paul Berry --- Local-branch: maint/next/remove_save_delegation lib/puppet/indirector.rb |2 +- lib/puppet/indirector/indirection.rb |2 +- lib/puppet/node/facts.rb |2

[Puppet-dev] [PATCH/puppet 2/5] Maint: Moved auto-signing logic into an indirector extension

Autosigning was previously accomplished by overriding CertificateRequest#save. This meant that it wouldn't work if certificate requests were saved via a direct call to Indirection#save. Changed it to use the indirector :extend mechanism, which works no matter how the save is invoked. Signed-off-b

[Puppet-dev] [PATCH/puppet 3/5] Maint: Add a default value for key in Facts::NodeExpirer#save

Changed to match the signature of the method being overridden. This will allow code to call Facts.indirection.save() without a key. Signed-off-by: Paul Berry --- Local-branch: maint/next/remove_save_delegation lib/puppet/node/facts.rb |2 +- 1 files changed, 1 insertions(+), 1 deletions(-)

[Puppet-dev] [PATCH/puppet 1/1] (#5422) Ship auth.conf as part of installing from source Add --configdir which defaults to /etc/puppet Add --no-configs to suppress config file installation. We're only

Signed-off-by: Nigel Kersten --- install.rb | 45 ++--- 1 files changed, 38 insertions(+), 7 deletions(-) diff --git a/install.rb b/install.rb index f7541c8..7627a8d 100755 --- a/install.rb +++ b/install.rb @@ -79,6 +79,7 @@ def glob(list) end # Set

[Puppet-dev] [PATCH/puppet 1/1] Bug #5423: This moves the home directory property before the uid property, thus minimizing room for damage when usermod is in use.

From: Jonathan Boyett Signed-off-by: James Turnbull --- lib/puppet/type/user.rb | 10 +- 1 files changed, 5 insertions(+), 5 deletions(-) diff --git a/lib/puppet/type/user.rb b/lib/puppet/type/user.rb index c8110bb..761d5d7 100755 --- a/lib/puppet/type/user.rb +++ b/lib/puppet/type/

Re: [Puppet-dev] Re: (#4974) Allow tags=main to apply all resources.

It's possible that I'm not understanding your suggestion, but: > Random idea. If stages themselves worked via tags instead of adding > something to the graph (and creating many-to-many dependencies), would > that be more functional? It would eliminate the main point of stages though; the whole po

Re: [Puppet-dev] Why is target a property and no param

On Nov 21, 2010, at 1:31 AM, Stefan Schulte wrote: > Hi, > > I have a general question about parameter vs property. I always thought > that a property is something that can be insync or outofsync, so thats > something that puppet can check and correct. > > But when I look at the current types, t

Re: [Puppet-dev] Re: (#4974) Allow tags=main to apply all resources.

I couldn't find a clean way of having them work via tags, and I think this model actually works better because it exposes that ordering directly in the graph, which I think is far easier to support in the long term. E.g., finding a cycle would be very hard if we were using tags, but is straight

Re: [Puppet-dev] Re: (#4974) Allow tags=main to apply all resources.

I agree, this isn't really the best way to do it, given stages. I think some kind of metatag like 'all' (which the tagmail report processor uses) or supporting '*' or something similar is the better move. On Nov 25, 2010, at 8:12 AM, Nigel Kersten wrote: > I'm not entirely convinced this is the

Re: [Puppet-dev] Re: (#4974) Allow tags=main to apply all resources.

Random idea. If stages themselves worked via tags instead of adding something to the graph (and creating many-to-many dependencies), would that be more functional? If you could use the tag() function to apply a tag to the class as normal, then define if that tag comes before or after the 'main' ta

Re: [Puppet-dev] [PATCH/puppet 1/2] Maint: Refactor tests to use .indirection.

On Tue, Nov 30, 2010 at 11:32 AM, Brice Figureau < brice-pup...@daysofwonder.com> wrote: > On 30/11/10 19:39, Paul Berry wrote: > > On Mon, Nov 29, 2010 at 10:47 PM, Luke Kanies > > wrote: > > > > What's the reason for this change? > > > > I assume it's related

Re: [Puppet-dev] [PATCH/puppet 1/2] Maint: Refactor tests to use .indirection.

On Tue, Nov 30, 2010 at 11:17 AM, Luke Kanies wrote: > On Nov 30, 2010, at 10:39 AM, Paul Berry wrote: > > On Mon, Nov 29, 2010 at 10:47 PM, Luke Kanies wrote: > >> What's the reason for this change? >> >> I assume it's related to the default route stuff, but I'd like to >> understand a bit more

Re: [Puppet-dev] Class ProviderAixuseradd is already defined in Puppet::Type::User

On Nov 30, 2010, at 12:04 AM, Héctor Rivas Gándara wrote: > On Mon, Nov 29, 2010 at 4:11 PM, Nigel Kersten wrote: >>> I am getting this error again, and it is anoying. I am using puppet 2.6.3. >>> >>> The problem appears randomly, I can not explain which changes makes it fail. >> >> Hector, are

Re: [Puppet-dev] [PATCH/puppet 1/2] Maint: Refactor tests to use .indirection.

On 30/11/10 19:39, Paul Berry wrote: > On Mon, Nov 29, 2010 at 10:47 PM, Luke Kanies > wrote: > > What's the reason for this change? > > I assume it's related to the default route stuff, but I'd like to > understand a bit more about this direction. In par

Re: [Puppet-dev] [PATCH/puppet 1/2] Maint: Refactor tests to use .indirection.

On Nov 30, 2010, at 10:39 AM, Paul Berry wrote: > On Mon, Nov 29, 2010 at 10:47 PM, Luke Kanies wrote: > What's the reason for this change? > > I assume it's related to the default route stuff, but I'd like to understand > a bit more about this direction. In particular, I thought the default r

[Puppet-dev] [PATCH/puppet 1/1] Maint: Make http handler code call the indirector through ".indirection"

In commit 71ecad9904c8c48c023e90e5fbea5b26b180c9cf we removed the delegation from model class to indirection for the "find", "search", "destroy", and "expire" methods. When we did this we neglected to modify http handler code to call the indirector directly. This patch makes the appropriate chang

[Puppet-dev] [PATCH/facter 2/2] (#4754) Change is_virtual logic to not enumerate virtual types

While looking at the patch for adding parallels to the virtual types David Schmitt noticed that it might be easier just to list the types that are NOT virtual since there's fewer of them. Paired-with: Nick Lewis Signed-off-by: Matt Robinson --- Local-branch: ticket/next/4754 lib/facter/virtual.

Re: [Puppet-dev] [PATCH/puppet 1/1] Maint: made upstart tests more robust.

I'm not sure whether Windows shell supports the "true" command, so you're right, this test may fail under Windows. As to your question about unit testing of Windows in general, this is currently a gap in our testing infrastructure. We don't have a windows box in our CI pipeline, and we know that

Re: [Puppet-dev] [PATCH/puppet 1/2] Maint: Refactor tests to use .indirection.

On Mon, Nov 29, 2010 at 10:47 PM, Luke Kanies wrote: > What's the reason for this change? > > I assume it's related to the default route stuff, but I'd like to > understand a bit more about this direction. In particular, I thought the > default route work would actually remove the find etc. meth

[Puppet-dev] [PATCH/facter 1/1] (#1423) Memory facts for Solaris

Add total memory from prtconf output, free from vmstat plus swap free and total from swap -l listing. Signed-off-by: Dominic Cleal --- Local-branch: tickets/master/1423 lib/facter/memory.rb | 51 ++-- lib/facter/util/memory.rb | 12 ++ 2 f

Re: [Puppet-dev] Class ProviderAixuseradd is already defined in Puppet::Type::User

On Mon, Nov 29, 2010 at 4:11 PM, Nigel Kersten wrote: >> I am getting this error again, and it is anoying. I am using puppet 2.6.3. >> >> The problem appears randomly, I can not explain which changes makes it fail. > > Hector, are you still having an issue? It's not clear whether your > resolution