On May 23, 2011, at 11:45 AM, Dan Bode wrote:
> This is an experimental set of patches that I wrote for composite namevars
> to work with parsedfile and to support purging.
>
> This is not ready to be merged, I would just like to get some input on the
> following:
> - does this work (for peopl
Hi all,
I've been thinking for a while of experimenting with graph databases --
especially Neo4j[1], but there are others out there -- and just this week I ran
across a graph language, Gremlin[2].
I know Volcane has done some experimentation with Neo4j, but has anyone else
messed with any of t
On Mon, May 23, 2011 at 4:01 PM, MolMorroco wrote:
> # 1. able to see confirm pluginsync = true works
> $ find /var/lib/puppet/lib/puppet/ -type f
> /var/lib/puppet/lib/puppet/type/custominstall.rb
> /var/lib/puppet/lib/puppet/provider/custominstall/custominstall.rb
>
> # 2. but I still get the fo
can you point me to a branch I can pull from?
On Mon, May 23, 2011 at 3:12 PM, Stefan Schulte <
stefan.schu...@taunusstein.net> wrote:
> On Mon, May 23, 2011 at 02:39:40PM -0700, Dan Bode wrote:
> > :name is special (and hardcoded) in several parts of the Puppet code.
> Could
> > you try using so
On May 23, 2011, at 11:53 AM, Jacob Helwig wrote:
> So, I'm working on #7188 and thought "hey, it would be great if we had
> built-in checks that these required parameters were actually specified".
> Especially since the current mount type claims that fstype is required,
> yet does nothing (other
On Mon, May 23, 2011 at 02:39:40PM -0700, Dan Bode wrote:
> :name is special (and hardcoded) in several parts of the Puppet code. Could
> you try using some other name that :name and see if it resolves those
> issues?
>
Doesnt resolve the issue. The test did pass before I applied your
patch. Here
:name is special (and hardcoded) in several parts of the Puppet code. Could
you try using some other name that :name and see if it resolves those
issues?
On Mon, May 23, 2011 at 2:25 PM, Stefan Schulte <
stefan.schu...@taunusstein.net> wrote:
> On Mon, May 23, 2011 at 11:45:45AM -0700, Dan Bode w
On Mon, May 23, 2011 at 11:45:45AM -0700, Dan Bode wrote:
> This is an experimental set of patches that I wrote for composite namevars
> to work with parsedfile and to support purging.
>
> This is not ready to be merged, I would just like to get some input on the
> following:
> - does this wor
On Mon, May 23, 2011 at 11:45 AM, Dan Bode wrote:
> For types with composite namevars, hash2resource
> should use namevar_join to populate tilte with
> namevar.
>
> Signed-off-by: Dan Bode
> ---
> Local-branch: issue/master/7629
> lib/puppet/type.rb |7 +--
> 1 files changed, 5 insertio
On Mon, May 23, 2011 at 11:45 AM, Dan Bode wrote:
> - for composite namevars to support purging
> and 'puppet resource' quiery all, it needs to
> be possible to map from the composite namevars
> to a single id.
>
> - the basic idea is that authors of types need to
> implement this method if t
On Mon, May 23, 2011 at 11:53:12AM -0700, Jacob Helwig wrote:
[...]
> Is there an implemented way of doing 'this parameter is required; fail
> at compile time'? Hopefully without having to do a defaultto of a bogus
> value, and having a validation failure against that.
You can specify a validate
On Mon, May 23, 2011 at 11:45 AM, Dan Bode wrote:
> - this ensures that resources are unique based on
> composite namevars.
> - I need to review why this is required..
>
> Signed-off-by: Dan Bode
> ---
> Local-branch: issue/master/7629
> lib/puppet/transaction.rb |4 +++-
> 1 files changed
Add a provider called mount that uses the output of mount to prefetch
current mounted devices and uses mount/umount to perform changes.
It may not work on any other System than Linux and even on this platform
it is not really well tested so far. Especially prefetching options and
the fstype may fa
Add a new type called mountpoint. This one should only ensure that a
device on a mountpoint is mounted/not mounted.
Valid properties are device, fstype and options but they are not
required because it's likely that there already is an fstab entry for
the described mount.
Mountoptions are currentl
As requested my attempt for a mountpoint type/provider for the splitted
mount type. The type should be usable but the provider is probably not
and doesnt have any tests yet.
One thought about a mountpoint provider in general:
How to handle mount options? I we specify options as parameters then
w
From: Rein Henrichs
ActiveRecord does not release connections back to its connection pool
automatically. This patch forces ActiveRecord to release its own
connections.
Signed-off-by: James Turnbull
---
Local-branch: tickets/2.7.x/3238
lib/puppet/rails.rb |2 +
lib/
Add specs to demonstrate that the instances method is currently broken
because it does not take the actual mountstate into account.
As a result running "puppet resource mount" on the commandline will
report every mount that appears in /etc/(v)fstab as unmounted.
Signed-off-by: Stefan Schulte
---
The instances method now behaves similar to the prefetch method. At first
/etc/(v)fstab is parsed to find mounts. After that the mount command is
run to update the ensure state from :unmounted to :mounted.
The instances method however does not take mounted devices into account
that are currently n
After discussion on redmine
http://projects.puppetlabs.com/issues/7300#note-10
here is the previous fix against 2.6.x which does not introduce the
change in behaviour:
* patch against 2.7.x (the previous one) will show mounts that are
specified in /etc/fstab AND mounted devices that do not app
its been like that for at least a year:
http://projects.puppetlabs.com/issues/4049
On Mon, May 23, 2011 at 11:53 AM, Jacob Helwig wrote:
> So, I'm working on #7188 and thought "hey, it would be great if we had
> built-in checks that these required parameters were actually specified".
> Especial
So, I'm working on #7188 and thought "hey, it would be great if we had
built-in checks that these required parameters were actually specified".
Especially since the current mount type claims that fstype is required,
yet does nothing (other than break things after a successful run) if
it's not speci
- Introduction of composite namevars places
extra restrictions on resource title format
(in order to guarantee uniqueness
it should be based on all namevars)
- Puppet was not explicitly handling the case
where the title is invalid
Signed-off-by: Dan Bode
---
Local-branch: issue/master/762
- previously, it was only matching based on the
name parameter. This caused a few problems.
- should not have to specify a name param
- it does not make as much sense with composite
namevars
- for composite namevars, it needs to match
records to resources based on all of the
namevars.
- it was previously hardcoded to use :name param
which does not work for composite namevars.
Signed-off-by: Dan Bode
---
Local-branch: issue/master/7629
lib/puppet/provider.rb |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/lib/puppet/provider.rb b/lib/puppet/provider
For types with composite namevars, hash2resource
should use namevar_join to populate tilte with
namevar.
Signed-off-by: Dan Bode
---
Local-branch: issue/master/7629
lib/puppet/type.rb |7 +--
1 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/lib/puppet/type.rb b/lib/puppet/t
This is an experimental set of patches that I wrote for composite namevars
to work with parsedfile and to support purging.
This is not ready to be merged, I would just like to get some input on the
following:
- does this work (for people who have pending composite namevar work)
- will this b
- this ensures that resources are unique based on
composite namevars.
- I need to review why this is required..
Signed-off-by: Dan Bode
---
Local-branch: issue/master/7629
lib/puppet/transaction.rb |4 +++-
1 files changed, 3 insertions(+), 1 deletions(-)
diff --git a/lib/puppet/transacti
- for composite namevars to support purging
and 'puppet resource' quiery all, it needs to
be possible to map from the composite namevars
to a single id.
- the basic idea is that authors of types need to
implement this method if they want to support
query all or purging.
Signed-off-by: D
Signed-off-by: Dan Bode
---
Local-branch: issue/master/7629
lib/puppet/type.rb |3 +--
1 files changed, 1 insertions(+), 2 deletions(-)
diff --git a/lib/puppet/type.rb b/lib/puppet/type.rb
index dc4764e..e62a764 100644
--- a/lib/puppet/type.rb
+++ b/lib/puppet/type.rb
@@ -893,8 +893,7 @@ cl
> Hm, does this work in case I have one keyattribute that is not "name"
> (like the fileresource where the keyattribute is "path")?
No, but neither does the existing code. :(
> If not, here is a second suggestion
>
> def self.title_join(hash)
> # extract all parameters that are keya
On Sun, May 22, 2011 at 05:56:29PM -0700, Dan Bode wrote:
> I just finished something very similar:
>
>
> def self.title_join(hash)
> "#{hash[:domain]}/#{hash[:type]}/#{hash[:item]}"
> end
>
> where the default case was:
>
> + def self.title_join(hash)
> +case key_attributes.length
Update changelog and adjust error message to provide some more
information to assist with debugging.
Write tests for the has_cf_class? method
Signed-off-by: R.I.Pienaar
---
Local-branch: bug/master/7627
lib/mcollective/util.rb |4 ++--
spec/spec_helper.rb |1 +
spec/unit/util_spec.r
From: Nicolas Szalay
Handle exceptions when trying to parse the classes file and log
a warning level message.
Signed-off-by: R.I.Pienaar
---
Local-branch: bug/master/7627
lib/mcollective/util.rb | 14 +-
1 files changed, 9 insertions(+), 5 deletions(-)
diff --git a/lib/mcollecti
Catch and raise again the SystemExit exception which allows the normal
exit method to be used instead.
Remove other uses of the exit! method in favor of the exit method
Adjust tests accordingly
Signed-off-by: R.I.Pienaar
---
Local-branch: bug/master/7626
lib/mcollective/application.rb |9 +
34 matches
Mail list logo
