I actually take the opposite approach. Modules from the forge go in a
directory named "forge_modules" and r10k will leave anything located in
modules alone.
From r10k.yaml:
forge 'http://forge.puppetlabs.com'
moduledir 'forge_modules'
On Friday, June 17, 2016 at 5:28:42 PM UTC-4, Rob
You'll need to install the package that provides augtool in order to use it.
yum provides augtool should show you what package to install.
On Friday, June 24, 2016 at 9:29:25 AM UTC-4, Shrawan Bhagwat wrote:
>
> Hi All,
>
> I have installed Augeas setup from yum repository using command yum
>
You may need to create a custom fact to determine the environment when
the agent runs. For example, here's a
custom fact that reads the environment from a file.
Facter.add("environment") do
setcode do
Facter::Util::Resolution.exec('/bin/cat /usr/local/etc/environment')
end
end
Depending on what you need you can install puppet packages directly from
the puppet repos as documented here.
https://docs.puppet.com/puppet/4.5/reference/install_linux.html#install-a-release-package-to-enable-puppet-collection-repositories
On Thursday, July 14, 2016 at 3:12:51 AM UTC-4,
I think I ran into a similar issue before. Try putting "eyaml" as the
first backend to see if that helps.
Here's a copy of our hiera.yaml file which works.
---
:backends:
- eyaml
- yaml
:hierarchy:
- "nodes/%{::trusted.certname}"
- common
:yaml:
# datadir is empty here, so hiera
It looks like you need to add a package_type for deb and a case
statement for Ubuntu. Would be a good time to patch the module and
submit a pull request. :)
On Friday, January 20, 2017 at 10:37:58 AM UTC-5, Salty Old Cowdawg wrote:
>
> Supposedly on the forge site it says the module was
puppet-dashboard is EOL and has not been updated in quite a while.
You'll need to move to a modern solution such as Puppetboard which does
work with Puppet Server and Puppetdb 4.
On Friday, February 10, 2017 at 3:06:23 PM UTC-5, Dimitri Yioulos wrote:
>
> Hi, All.
>
> Previously, I had a
ensure => '5.0' should work but you may want to add a dependency to the
package resource to ensure that the file is downloaded first.
package { "package":
ensure => '5.0',
provider => 'rpm',
source => '/tmp/package-5.0.el6.noarch.rpm',
require =>
Looks like your master's host name isn't included on the certificate. Are
you able to run the agent using puppet agent -t --server puppet ?
You can also see what host names are returned by the master's cert using
openssl.
openssl s_client -connect puppet:8140
This should show you the entire
There seems to be a bug in the mount provider with the new puppet-agent
1.8 package. I have a node that is reporting changed file system
options every time that puppet runs although nothing is actually
changing. I downgraded back to puppet-agent 1.7.1 and this behavior
goes away.
Puppet keeps a cache of classes and resources on every node. You can
look at the /opt/puppetlabs/puppet/cache/state/resources.txt file to see
which resources are managed by puppet.
On Tuesday, October 11, 2016 at 7:41:11 AM UTC-4, Al wrote:
>
> Hi, subj.
> Configuration is very complex,
%x is a ruby method which captures command output. IMO you can do most of
what you need using native ruby methods, there's no need to pipe output to
grep/sed/awk since ruby has built in pattern matching functions and if
you're using ruby you might as well do it the ruby way.
For example, this
You can do everything you need entirely in ruby. For example, here's a
custom fact that returns the number of screens reported by the X server.
Facter.add(:screen_count) do
confine :kernel => 'Linux'
setcode do
@screens = Facter::Core::Execution.exec("/usr/bin/xrandr -display
This is pretty much exactly what module data is for. Check
out https://docs.puppet.com/puppet/4.9/lookup_quick_module.html for more
details.
On Friday, September 30, 2011 at 2:33:50 PM UTC-4, Jeff Falgout wrote:
>
> We're in the situation of dealing with multiple operating systems (and
>
notify events are queued and grouped together so nginx should only be
restarted once when puppet runs. You may also need to include nginx in
your profile so that the service is properly defined.
On Friday, March 3, 2017 at 7:00:36 AM UTC-5, Steve Button wrote:
>
> Hi,
>
> I've got a small
Also, it looks like somebody tried to fix this previously but the changes
weren't merged in.
See https://github.com/voxpupuli/puppet-nginx/issues/333 for example. IMO
there should be a class parameter to control this behavior so you might
want to submit a PR with any changes you make.
On
This would be the ideal but you *can* use the rpm provider when needed.
For example:
package { 'jdk':
ensure => installed,
provider => 'rpm',
source => '/pub/oracle/jdk-8-linux-x64.rpm',
}
This will install the rpm using the defined source path. In our
environment the /pub
You do not *have* to quote strings however I have ran into issues with hiera
turning strings that look like integers into an integer which causes problems
when you attempt to do a key lookup using a string value or pass a value to a
parameter that expects a string value.
For example, there
I usually do this within the manifest itself. For example:
if $facts['osfamily']!='RedHat'{fail("$operatingsystem is not supported")}
On Tuesday, August 22, 2017 at 7:34:12 AM UTC-4, plapp...@gmail.com wrote:
>
> Good morning :)
>
> I'm currently writing a puppet module for the OpenBSD
Been through a similar upgrade myself. The first step would be to spin up
a new puppet master running Puppet Server. You can copy over the SSL dir
from your old/current master to avoid SSL errors on the agents. For
testing you'll want to make sure your manifests work correctly using the
Hello,
Does anybody know how to make rspec-puppet work with modules that use
custom types for class parameters? I am getting an error when I run spec
tests such as follows.
# Puppet::ParseError:
# Expected parameter 'ensure' of 'Class[Logrotate]' to have type
Logrotate::Ensurable,
Thanks. I'll create an issue with the rspec-puppet project.
On Tuesday, October 24, 2017 at 10:15:14 AM UTC-4, Henrik Lindberg wrote:
>
> On 23/10/17 22:29, Michael Watters wrote:
> > Hello,
> >
> > Does anybody know how to make rspec-puppet work with modules th
Not familiar with this module but you could set up a custom exec resource to
restart the network when $facts['ipaddress'] does not match what is defined in
hiera. We do something similar for SELinux to make sure the server is actually
in the mode that it should be.
On Thursday, May 24,
> Is there in fact any reason for concern that applying class autofs before
applying one or both of the other two might result in a Puppet failure or a
target system misconfiguration?
In this case, yes. In order for autofs to work properly NFS must be
configured first and in order for NFS to
the issue but it still feels like a
kludge instead of a proper solution.
On Thursday, August 9, 2018 at 3:33:47 PM UTC-4, Ben Ford wrote:
>
> Hi Michael. Can you share the code you tried and the errors you ran into?
>
> On Thu, Aug 9, 2018 at 11:30 AM Michael Watters > wrote
Also, does the release-archives site have rsync enabled?
On Thursday, August 9, 2018 at 7:29:40 PM UTC-4, Rob Kenefeck wrote:
>
> Hi Eric, when will this be happening?
--
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To unsubscribe from this
Hello,
Is it possible to define metaparameters for a resource in Hiera? For
example, I have a class that is currently declared within another class as
follows.
if !defined(Class['autofs']) {
class { 'autofs':
require => [
Class['authconfig'],
What happens when there's a bug fix or security update that you need?
Pinning versions also only works as long as you have no *other* modules
that depend on a newer version of the apache module.
IMO the apache module should not be dropping support for an Apache release
that is still widely
It sounds like your atqapache::vhost type is attempting to create
the /var/www/dev/user1/client1 directory for every vhost that is defined.
Can you post your definition for this type here?
On Friday, September 7, 2018 at 5:05:44 PM UTC-4, Ugo Bellavance wrote:
>
> Hi,
>
> I have made a
ption but this way ensures that the resource is only
created once.
On Monday, September 10, 2018 at 10:22:18 AM UTC-4, Ugo Bellavance wrote:
>
>
>
> On Monday, September 10, 2018 at 9:46:18 AM UTC-4, Michael Watters wrote:
>>
>> It sounds like your atqapache::vhost type is attem
the package resource. Then as the file
> doesnt exist the run fails.
> Back to drawing board.
>
> Appreciate the quick reply.
>
>
> On Tuesday, September 11, 2018 at 7:27:49 AM UTC-5, Michael Watters wrote:
>>
>> What you have there looks fine but I would change
What you have there looks fine but I would change the elsif to an else
statement.
On Tuesday, September 11, 2018 at 8:08:53 AM UTC-4, Jason McMahan wrote:
>
> Good day,
> I am attempting to add require to an if statement and not sure i am doing
> it right. Any help would be greatly
Most likely there are dependencies which would also need to be updated and
Redhat likes to avoid breaking things in a RHEL release. If you really
need the latest version of puppet installed the packages from the official
repos work fine.
On Thursday, April 5, 2018 at 12:26:19 PM UTC-4,
Puppetdb data is all stored in postgresql so you should be able to copy the
reports table from one server to the other. For example, run this on your
prod node.
pg_dump -h puppetdb-dev -U puppetdb -W -d puppetdb -t reports | psql -U
puppetdb puppetdb
I'm not sure if you need more than just
Thanks. I found that adding "require 'parseconfig'" to the
spec/spec_helper.rb file also allows the tests to run. bundle exec is
failing with dependency errors as well.
On Monday, April 23, 2018 at 5:24:08 PM UTC-4, Garrett Honeycutt wrote:
>
> On 4/23/18 4:38 PM, Micha
Use the puppetlabs-mysql module to manage the grants for each user/db. You
can have different values for each environment stored in hiera.
On Friday, April 20, 2018 at 11:42:22 AM UTC-4, Jonathan Preston wrote:
>
> I've searched for answers, but couldn't find anything quite matching my
> use
Has anybody used rspec-puppet to run tests using a function that requires
external ruby gems? I'm running unit tests on resources that call
functions in the puppet_thycotic module however rspec is failing with an
error as shown below.
Failure/Error: require 'parseconfig'
I've done this for a few nodes but I'm not sure how this would be an
improvement over just enabling autosign. Private keys should remain
private to a node and should never be transmitted over the network if
possible.
On Wednesday, March 28, 2018 at 3:10:35 PM UTC-4, Eric Sorenson wrote:
>
>
I'm having an issue with puppet repeatedly attempting to mount resources
which are configured as bind mounts on one of my nodes. Here is a copy of
the relevant lines from my manifest.
$koji_fedora_releases = ['28', '29', '30']
$koji_fedora_releases.each |$release| {
['released',
Hello,
I see the puppetlabs-release-pc1 package is available on yum.puppetlabs.com
however it doesn't actually show as an available update when running "yum
update".
For example, here's the output on one of our servers.
```
[root@server1 yum.repos.d]# yum update puppetlabs-release-pc1
Loaded
How does this change affect private mirrors? Will rsync services be
available on the archive repos?
On Monday, April 22, 2019 at 5:31:09 PM UTC-4, Molly Waggett wrote:
>
> TLDR: On May 14, 1. yum release packages will move to the top-level of
> yum.puppet.com; 2. links to nightly release
; ensure => 'mounted',
> device => '/home',
> dump=> '0',
> fstype => 'none',
> options => 'rw,bind,seclabel,relatime,errors=remount-ro',
> pass=> '0',
> target => '/etc/fstab',
> }
>
>
> -Original message-
> *From:
I'm having an issue with the puppetlabs-stdlib module causing compilation
failures using the puppet apply command.
> Error: Class 'stdlib' is already defined (file:
/mnt/live/puppeteer/conf.d/puppet/modules/stdlib/manifests/init.pp, line:
16); cannot redefine (file:
e packages from
> http://release-archives.puppet.com/yum/ if you want to continue using the
> PC1 repository. Neither of these repos will receive future updates.
>
> If you have additional questions, please let me know!
>
> On Wed, Jun 5, 2019 at 10:36 AM Michael Watters
Thanks. Unfortunately that doesn't seem to be the issue. There is no file
named .resource_types within the puppet configuration dir.
>
> You may be running into
> https://tickets.puppetlabs.com/browse/PUP-9602. Try deleting
> /mnt/live/puppeteer/conf.d/puppet/.resource_types
>
> Josh
>
>
Are there any plans to package puppet for SLES 15? I only see packages for
SLES 11 and 12 at http://release-archives.puppet.com/yum/sles.
--
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To unsubscribe from this group and stop receiving emails
already as well.
[image: Screenshot_SLES15_2019-07-02_11:41:26.png]
On Tuesday, July 2, 2019 at 11:35:47 AM UTC-4, Michael Watters wrote:
>
> Are there any plans to package puppet for SLES 15? I only see packages
> for SLES 11 and 12 at http://release-archives.puppet.com/yum/sles.
&
There appears to be an issue with the puppet agent on Fedora 33 RC1.
Running the agent against a test environment results in errors as follows.
Does anybody know how to resolve this?
Info: Using configured environment 'dev27_f33'
Info: Retrieving pluginfacts
Error:
Why do these feel like homework questions? :D
You can add logic to your manifest based on what environment the agent is
running under using simple if statements. For example:
if $facts['environment'] != 'production' {
notify { 'This is a non-production node!': }
}
If you're not
ram files/telegraf',
refreshonly => true,
}
On Tuesday, December 8, 2020 at 1:29:55 PM UTC-5 Michael Watters wrote:
> I tried that as well and it simply results in an error stating "Could not
> find command: c:/program".
>
> On Monday, December 7, 2020 at 2:06:53
e --service install' :
> ...
> }
>
> > On 7. Dec 2020, at 18:17, Michael Watters wrote:
> >
> > I have an exec resource in a manifest for our Windows nodes however the
> command is failing each time that puppet runs. Here is the definition of
> the reso
Probably need to quote the executable due to spaces:
exec { '"c:/program files/telegraf/telegraf.exe" --service install':
Looks like that worked. Thanks for the help.
On Tuesday, December 8, 2020 at 1:58:12 PM UTC-5 Josh Cooper wrote:
> On Tue, Dec 8, 2020 at 10:37 AM Michael W
52 matches
Mail list logo