L --
I unfortunately don't currently have access to the source, but how
does validation on the overrides work with a source here?
That source should be used during the validation process, where a
resource confirms that a new parameter value is from a subclass of the
original parameter
On 3 August 2010 23:03, Rein Henrichs r...@puppetlabs.com wrote:
From: Rein Henrichs re...@reinh.com
Rewrite of uptime facts and supporting utility methods. Works on unix,
BSD, windows. No longer makes redundant system calls.
Uses Facter::Util::Uptime utility methods:
Thanks for this, not
Hi Jesse,
Thanks for fixing the typo in the patch.
I'm still not liking how we/I fixed this problem.
Did you had a look to the potential issues of caching those module
directories?
Did you experience compilation speed-up with the patch?
Also, I'm thinking that this module_directories function
On Wed, 2010-08-04 at 23:34 +1000, Nicolas Brisac wrote:
I'm getting the same error with a new custom fact in module/lib.
The puppetmaster is running on Apache2 + Passenger (setup works fine
in production with 0.24.8).
With Puppet 2.6.0 (Debian package) the file is created empty.
With
On Aug 4, 2010, at 6:18 AM, Brice Figureau wrote:
Hi Jesse,
Thanks for fixing the typo in the patch.
I'm still not liking how we/I fixed this problem.
Did you had a look to the potential issues of caching those module
directories?
Did you experience compilation speed-up with the patch?
I
Brice --
If that doesn't pan out could the problem be with our format_to_mime
(i.e. the content type, rather than the content length)? That seems
equally odd, but it would explain why there isn't widespread
discussion of the problem (which I would expect if passenger were
requiring content
On Wed, 2010-08-04 at 08:03 -0700, Luke Kanies wrote:
On Aug 4, 2010, at 6:18 AM, Brice Figureau wrote:
Hi Jesse,
Thanks for fixing the typo in the patch.
I'm still not liking how we/I fixed this problem.
Did you had a look to the potential issues of caching those module
B ( L) --
*laugh* We need to set up a web cam. To summarize the discussion in
the bull pen yesterday:
I'm still not liking how we/I fixed this problem.
As a permanent solution (2.7.x), agreed. As a hot fix (2.6.1), we
like it better than any of the ready alternatives. It isn't elegant,
but
On Aug 4, 2010, at 8:35 AM, Brice Figureau wrote:
On Wed, 2010-08-04 at 08:03 -0700, Luke Kanies wrote:
On Aug 4, 2010, at 6:18 AM, Brice Figureau wrote:
Hi Jesse,
Thanks for fixing the typo in the patch.
I'm still not liking how we/I fixed this problem.
Did you had a look to the potential
On Wed, Aug 4, 2010 at 5:02 AM, Paul Nasrat pnas...@gmail.com wrote:
+def self.uptime_proc_uptime
+if File.exists? uptime_file
+r = File.read uptime_file
+r.split( ).first.to_i
+end
+end
snip
end
-def self.get_uptime
-
On Aug 4, 2010, at 8:41 AM, Markus Roberts wrote:
[...]
Did you experience compilation speed-up with the patch?
I haven't tested this patch, but it basically re-adds caching that
was
present in 0.25, and I added it there because we saw exactly what
we're
seeing here - a huge amount of
On Wed, 2010-08-04 at 08:26 -0700, Markus Roberts wrote:
Brice --
If that doesn't pan out could the problem be with our format_to_mime
(i.e. the content type, rather than the content length)?
This returns format.mime (format being the current renderer), which I
hope is a string or we'll
Brice --
Agreed on almost everything with one niggling doubt.
Which is coherent with the error message. If we ask users the size of
the file it chokes on, I'm sure will find that it matches the fixnum we
find in the exception.
You are probably correct but it looks as if the error message:
L --
...my guess is
that it might actually make more sense to cache at a higher level - if the
environments or modules track whether a given name (type, function, etc.)
was already loaded or failed to get loaded, then there's no need to do any
file scanning.
Paul's doing the preliminary
Excerpts from Paul Berry's message of Wed Aug 04 08:43:54 -0700 2010:
On Wed, Aug 4, 2010 at 5:02 AM, Paul Nasrat pnas...@gmail.com wrote:
You've moved to read from cat, that'll likely break on RHEL 4 see
Correct me if I'm wrong, but it looks like the core problem is that reading
from /proc
When generating the error message when we can't find any instances for a
search request, we were inspecting the request which now contains
an environment instance which itself contains a lots of things (including
all the known resource types).
Thus it was generating a very large reason (ie the
+1
On Wed, Aug 4, 2010 at 1:35 PM, Brice Figureau
brice-pup...@daysofwonder.com wrote:
When generating the error message when we can't find any instances for a
search request, we were inspecting the request which now contains
an environment instance which itself contains a lots of things
Rethinking my +1: is the right place to fix this
lib/puppet/indirector/request.rb's to_hash (replacing the environment
with its name) to catch they other analogous cases?
-- Markus
---
The power of accurate observation is
commonly called
A spec was failing since it wasn't updated when the #4364 change was
made.
Signed-off-by: Jesse Wolfe jes5...@gmail.com
---
spec/unit/transaction/resource_harness_spec.rb |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/spec/unit/transaction/resource_harness_spec.rb
On 04/08/10 22:51, Markus Roberts wrote:
Rethinking my +1: is the right place to fix this
lib/puppet/indirector/request.rb's to_hash (replacing the environment
with its name) to catch they other analogous cases?
I thought about this, but I guess I was in too lazy.
Apparently to_hash is never
Signed-off-by: James Turnbull ja...@lovedthanlost.net
---
conf/solaris/smf/svc-puppetd |2 +-
conf/solaris/smf/svc-puppetmasterd |2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/conf/solaris/smf/svc-puppetd b/conf/solaris/smf/svc-puppetd
index 9036b50..b6cf057
Yes I can confirm that:
This is the data that it sent: [Content-Type]
*** Exception NoMethodError in
PhusionPassenger::Rack::ApplicationSpawner (undefined method `each'
for 1268 :Fixnum) (process 28465):
Which is coherent with the error message. If we ask users the size of
the file it
* Remove hard-coded facts terminus in master
* Change facts_terminus default to 'yaml' for master and 'facter' for
everything else.
Paired-with: Matt Robinson m...@puppetlabs.com
Signed-off-by: Rein Henrichs r...@puppetlabs.com
---
lib/puppet/application/master.rb |3 ---
+1, though I wish there were a non-heinous way to test that.
On Wed, 2010-08-04 at 17:21 -0700, Rein Henrichs wrote:
* Remove hard-coded facts terminus in master
* Change facts_terminus default to 'yaml' for master and 'facter' for
everything else.
Paired-with: Matt Robinson
On Aug 4, 2010, at 5:21 PM, Rein Henrichs wrote:
* Remove hard-coded facts terminus in master
* Change facts_terminus default to 'yaml' for master and 'facter' for
everything else.
Paired-with: Matt Robinson m...@puppetlabs.com
Signed-off-by: Rein Henrichs r...@puppetlabs.com
---
Excerpts from Luke Kanies's message of Wed Aug 04 17:42:55 -0700 2010:
On Aug 4, 2010, at 5:21 PM, Rein Henrichs wrote:
The way this has generally been done is that a given application will
override this value if appropriate, which makes this a bit cleaner, IMO.
E.g., have the 'master'
The patch for #3904 should not have been accepted; it's buggy and the use case
(supporting having nodes in ldap more than once with the same name but distinct
records) is ill-conceived.
This commit reverts the patch (a7884b47) and the previous attempt to fix it
(e6709da4), restoring the old
27 matches
Mail list logo
