It's not uncommon to run a small staging environment for just this
purpose. Have you evaluated something like that and found it wanting?
On Fri, Mar 2, 2012 at 7:26 PM, Ryan Bowlby wrote:
> Hi All,
>
> We'd like to do some form of testing of our module changes against
> production nodes before be
I did up a nifty deployment engine using Jenkins. Give the devs/CM a form
(e.g. "silo", application versions, etc). It would figure out what it
needed to deploy and then do so, complete with telling the Nagios system to
disable checks while everything was going on. Foreman/Puppet could be the
ri
Hi All,
We'd like to do some form of testing of our module changes against
production nodes before being released into production. While somewhat
expensive it seems doing a noop against all nodes using the modified
module is the best way to determine unexpected results.
The question then is how a
The gem includes everything you will need... the (basic) difference is
simply running "puppet agent" versus "puppet master" -- some packages just
hide the simplicity (probably a legacy thing?).
Note: I do not believe things like the sysconfig, logrotate and startup
scripts get installed by default
Any chance you copied over a bare metal box image, or something?
In any case, facter would seem to have only a handful of checks to
determine "virtual" or non-virtual... unfortunately I don't have a vmware
box in front of me to verify this, but you should be able to find facter's
"virtual" tests i
I can only find the puppet gem. Isn't there a puppet-server gem? I have a
working environment without rails. The following is installed on my puppet
master:
puppet-2.6.7-1.el5
puppet-server-2.6.7-1.el5
And the following is installed on my clients:
puppet-2.6.7-1.el5
You mention only one gem.
I'm not sure I understand his setup, or what he means by "minimal
install".
My environment on the VMware image is CentOS 5.7, it is a full release
and the NFS mount contains a full release of Puppet and Ruby 1.8.x.
Perhaps there's something that Facter gets wrong when it's being
called from a non
Potentially related?
http://projects.puppetlabs.com/issues/10232
On Fri, Mar 2, 2012 at 2:44 PM, Forrie wrote:
> I have Puppet mounted to a couple of machines via NFS (along with
> Ruby) -- and it works fine. Except, I just noticed that it fools
> Facter in to believing it's a physical machi
On Mar 2, 2012, at 3:43 PM, rjl wrote:
> Thanks Michael,
> I have a working environment using ruby 1.8.5 (The Centos 5 default).
> However, I want to change my puppet framework to be database driven. For
> that, I would like to use rails activerecord. Now the problem is that rails
> is not su
I have Puppet mounted to a couple of machines via NFS (along with
Ruby) -- and it works fine. Except, I just noticed that it fools
Facter in to believing it's a physical machine, when in fact one of
them is a VMware host.
We don't really use this setting, but I'm concerned other settings
might n
Thanks Michael,
I have a working environment using ruby 1.8.5 (The Centos 5 default).
However, I want to change my puppet framework to be database driven. For
that, I would like to use rails activerecord. Now the problem is that
rails is not supported in 1.8.5.
Any suggestions would be greatly
I just dealt with something similar regarding installing puppet agent vs.
master and whether mcollective client (and thus activemq) should be
installed. However, I'm including my base class rather than inheriting it.
So how about something like this?
# templates.pp (imported into site.pp)
class sy
It actually require ruby(abi) = 1.8
This is likely only satisfied by Ruby 1.8.5 in your repositories.
This is the way the rpm package is currently built. This is in the
problem of changing, as 1.9.3 is now becoming the default in Fedora
17. I suspect a spec version very soon will have a this up
Hi,
No, puppet won't install and configure itself.
However you can provision puppet onto your cluster at build time and configure
it at that time. We use cobbler to provision machines with the required
software and initial configuration. Foreman is another option here.
On an existing small clu
I have compiled and installed ruby 1.9.3 on Centos 5
# ruby -v
ruby 1.9.3p125 (2012-02-16 revision 34643) [x86_64-linux]
Now I am trying to install puppet 2.7.11. However, it is requiring that
ruby version 1.8.5-5.el5_4.8.
# yum install puppet
Installing:
puppet
Have you setup Dashboard as an ENC?
http://docs.puppetlabs.com/dashboard/manual/1.2/bootstrapping.html
On Fri, Mar 2, 2012 at 7:55 AM, Peter Berghold wrote:
> Is there something on the puppet master or puppet agent that needs to be
> tweaked for class assignments to be picked up form dasbhboard w
I'm using Puppet as a configuration management tool in the open source
Collabograte project that I recently announced:
http://kartiksubbarao.com/announcing-collabograte
Currently, the puppet manifests enable integration points among the
following software packages: Cyrus IMAP, ejabberd, INN, Me
We do something similar to what you describe with foreman (which can
be used as an ENC). The user sets the my-app-version parameter on the
node, then runs puppet on that node. The main drawback is that
foreman does not currently have a permission for puppetrun, so the
users have to be admins, so
On Fri, Mar 2, 2012 at 08:56, Romeo Theriault wrote:
> Hi, I'm just getting started with puppet and am looking for some best
> practices on how to handle node and module inheritance issues. I'm
> planning to start using heira so want to plan my implementation around
> hiera specifics.
>
> Specific
On Fri, Mar 2, 2012 at 10:37 AM, Peter Berghold wrote:
>
> In Perl we have the functionality of
>
> if ( defined($somevar) ) {
> ... do something
> }
>
if ( $somevar != undef) {
...
}
You can test this quickly using the --execute flag to puppet apply:
% puppet apply -e 'if $foo == un
On Fri, Mar 2, 2012 at 1:45 PM, Peter Berghold wrote:
> My puppet master server has a fqdn of chidpmaster.some.domain.tld with a
> cname pointing to it of puppet.domain.tld.
>
>
Figured this one out on my own. Seems in my site.pp file I have:
filebucket { "main":
server=>"chidpmaster.some
Thanks for the suggested method, I will put up a pilot environment and
see if this solution works for me!
On Feb 28, 2012, at 9:41 PM, Brian Gallew wrote:
Allow me to offer a couple of alternatives:
1) If Puppet is not otherwise doing stuff with the domains, then
stop trying to manage the b
Hi, I'm just getting started with puppet and am looking for some best
practices on how to handle node and module inheritance issues. I'm
planning to start using heira so want to plan my implementation around
hiera specifics.
Specifically, one item I can't seem to find a clean way of dealing
with i
I actually have figured this out (with some help from the list.) Seems I
was encountering a PEBKAC error. :-)
I hadn't set up puppet to use dashboard for external node classification.
(so that's what it's useful for! hmmm...)
>
--
Peter L. Berghold
Owner, Shark River Technical Solutions LLC
My puppet master server has a fqdn of chidpmaster.some.domain.tld with a
cname pointing to it of puppet.domain.tld.
For the first time since I started playing around with puppet I'm now
seeing an error when applying a template --
err: /Stage[main]/Sudo/File[sudoers]/content: change from
{md5}c5da
On Fri, Mar 2, 2012 at 04:12, Peter Berghold wrote:
>
>
> On Fri, Mar 2, 2012 at 12:55 AM, Romeo Theriault
> wrote:
>>
>>
>> In the Dashboard, when they say "classes" they really mean "module".
>>
>
>
>
> This begs a follow-on question:
>
> if classes == modules in Dashboard then maybe I need to
In Perl we have the functionality of
if ( defined($somevar) ) {
... do something
}
Is there an equivalent you can use within a puppet manifest?
--
Peter L. Berghold
Owner, Shark River Technical Solutions LLC
--
You received this message because you are subscribed to the Google Grou
Ok i did some reading and watched a youtube clip on it also.
Heira can be set in the params manifest which is what ll do - quite
dangerous that ill be using this approach at a company i work for :-)
Next i have established my hierarachy as such :
hierarchy: - %{_domain}/%{_environment}/%{_cluster
I have several hundred nodes I'm migrating to being managed by Puppet. We
do have a handful of older servers that can't run Puppet (RH8 etc), and
won't be upgraded any time soon, unfortunately. I'd like to be able to have
an 'artificial' node created with accurate inventory data displayed in
Pu
On Fri, Mar 2, 2012 at 11:09 AM, Kenneth Lo wrote:
> Texternal_nodes parameter to point to a process that returns a yaml
> formatted "node catalogs". If your ENC configuration is correct (in this
> case dashboard), puppet would do the rest.
>
>
That was it!
I think I am going to take the time
The way I understand it is that, if your puppet installation utilize ENC, you
simply setup the puppet.conf's [master] external_nodes parameter to point to a
process that returns a yaml formatted "node catalogs". If your ENC
configuration is correct (in this case dashboard), puppet would do the r
Is there something on the puppet master or puppet agent that needs to be
tweaked for class assignments to be picked up form dasbhboard when a puppet
agent runs?
wondering if I missed something...
--
Peter L. Berghold
Owner, Shark River Technical Solutions LLC
--
You received this message beca
Furthermore I can verify the cert client side w/ the ca:
root::wave { 10:34:20 Fri Mar 02 }
~-> openssl verify -CAfile /var/lib/puppet/ssl/certs/ca.pem
/var/lib/puppet/ssl/certs/wave.pem
/var/lib/puppet/ssl/certs/wave.pem: OK
?
On Fri, Mar 2, 2012 at 10:14 AM, Matthew Nicholson
wrote:
> So,I m
So,I made a stupid move this morning I'm trying to correct. While
trying to getthe puppet master to config itself, i moved its certs
sideways, and regenerated. Durring this time i did a ntp sync and
found i was about 2 seconds off. This little test failed and I
decided I had a better way to do it(
On Fri, Mar 2, 2012 at 12:55 AM, Romeo Theriault
wrote:
>
> In the Dashboard, when they say "classes" they really mean "module".
>
>
This begs a follow-on question:
if classes == modules in Dashboard then maybe I need to take a second look
here.
The module that I am writing from scratch to wor
Hi,
This is the third or fourth time this happens. But puppetd gets a zombie
shell childprocess and then never finishes the run.
/opt/tc-puppet/bin/ruby /opt/tc-puppet/sbin/puppetd
\_ [sh]
How do I begin looking on what can be wrong?
This error has appeared on both 2.7.6 and 2.7.11.
Regards,
Based on everything I have read thus far and my experiences working with
Ruby 1.9.{2,3} and Puppet 2.7.x, you should stick with Ruby 1.8.7 until
later versions of Puppet 2.7.x/2.8. IMHO
On Thursday, March 1, 2012 2:40:08 PM UTC-5, Chad Huneycutt wrote:
>
> So does anyone have a working setup wit
We haven't actually done this in production yet, but we've discussed
it quite a bit. Our current theory for things like this is:
1) MySQL-based External Node Classifier. Developers get
(authenticated, ACL'ed) access to a simple PHP script with two
options: a dropdown list of modules for their app
I've not but it is a good idea :)
thanks
On 2 March 2012 11:23, denmat wrote:
> HI,
>
> have you tried including the mounts module in your application.pp?
>
> include mounts::mount1
>
> Either there or in your node definition should do it.
>
> Cheers,
> Den
>
> On Fri, Mar 2, 2012 at 9:39 PM, Pe
Hi All,
I am very new to Puppet. Need some help regarding following.
Lets say we have a cluster of 20 machines which i want to manage
through Puppet.
Do I need to deploy Puppet itself manually on each machine?
OR
Is there a way that i install Puppet Master and ask it to deploy the
puppet on remai
Hi Folks
To begin with i am looking to deploy a completely new puppet config
based on the truth module
https://github.com/jordansissel/puppet-examples/tree/master/nodeless-puppet/modules/truth
However i want a way of separating data from modules based on this
hierarchy:
domain - ie blogs.com =>
I have a template for zabbix_agentd.conf.erb that has the following line in
it:
<% if has_variable?( "mysqlsrv" ) then %>
UserParameter=FromDual.MySQL.check,/usr/local/mysql_performance_monitor/FromDualMySQLagent.pl
/usr/local/mysql_performance_monitor/etc/FromDualMySQLagent.conf <% end %>
I
Hi,
For deployment we do not usually use puppet. The deployment we do with are
puppet are for stable in house packages.
This is then done by releasing a new version in our package environment and
utilizing ensure => latest for the package type.
But for frequent deployment methods I would perso
HI,
have you tried including the mounts module in your application.pp?
include mounts::mount1
Either there or in your node definition should do it.
Cheers,
Den
On Fri, Mar 2, 2012 at 9:39 PM, Peter Horvath wrote:
> Hello,
>
> I have 2 modules and 1 modul has a resource type in a class which
>
Hi Krzysztof,
>> I guess the question boils down to: Is there a way to cast from Fixnum
>> to String?
>
> Yes, look below :)
Excellent!
> Take a look on this:
>
> https://github.com/kwilczynski/puppet-functions/blob/master/lib/puppet/parser/functions/num2str.rb
>
> And this:
>
> https://github.c
Here is an example using inline templates. Should describe how you can do it.
class myclass {
}
class myclass::config {
$myvar = "this is class text"
}
class myclass::template {
include myclass::config
$myvar = $myclass::config::myvar
$mytemplate = inline_template("
Here is some text.
Hello,
I have 2 modules and 1 modul has a resource type in a class which
would require an other class from and other module:
module/mounts/manifest/mount1.pp
module/server/manifres/applicaton.pp
Application creates a directory under /srv for itself and /srv should
be mounted from the shared stora
Hi
I'm in the process of looking for a way to have developers deploying
on their test systems without intervention of sysadmins, to solve this
i'd like to use Puppet (either the OSS version or Enterprise,
whichever solves the problem).
I can manage to only grant access to certain systems and limi
48 matches
Mail list logo
