That ticket looks like the issue I'm having!
I look forward to the upcoming releases so we can start working on a plan
to upgrade. If the new puppetserver would talk nicely to older puppet's,
that would be great. It's going to suck when we have to bring up systems
that have been offline and h
On Fri, Apr 17, 2015 at 3:12 PM, Ken Bowley wrote:
> Puppet 4 needs a lot of new directories in order to even attempt to run.
> It would be nice if puppet had a way to create the directory structure that
> it needs in order to run.
>
Hi Ken,
I've just filed https://tickets.puppetlabs.com/browse
Puppet 4 needs a lot of new directories in order to even attempt to run.
It would be nice if puppet had a way to create the directory structure
that it needs in order to run.
For now I'm trying to ignore the seemingly non-existent upgrade path from
puppet 3 to puppet 4 that doesn't involve fir
Ken Bowley writes:
> We've been using Puppet since the 0.24 days, and it's always been fairly
> simple to add in a new system.
> Do a base install (Linux or Mac OS X)
> Install Ruby if Linux
> gem install puppet
> run puppet
> sign the cert on the puppetmaster
> let puppet take care of everythi
We've been using Puppet since the 0.24 days, and it's always been fairly
simple to add in a new system.
Do a base install (Linux or Mac OS X)
Install Ruby if Linux
gem install puppet
run puppet
sign the cert on the puppetmaster
let puppet take care of everything else.
This was easy, easy to docu
Hi,
we current use the rabbitmq module in a production environment and for our
use case we need to disable explicit the ssl port for the management
console but enable it for the RabbitMQ self.
This is needed because the web interfaces are using one SSL certificate
that will only installed on th
I did just that, used a template to populate some external facts:
[root@host123cluster123 ~]# cat /etc/facter/facts.d/inventory.txt
cluster=123
servertype=typename
stype=typename
datacenter=dcname
owner=groupname
[puppet exp] $ cat inventory.erb
<%# these will be blank if the variable does not
Is it possible to reference a top scope variable passed in from an ENC and
set a fact based on that? That is,
parameters:
var: value
I know it's the other way around that top scope variables are set from
facts, but just trying to figure out the best way to do this other that
writing out a scri
Hello,
i am setting up my new puppet env, i am using "Katello" for HostGroups and
applying classes to hostgroups ... and using Hiera for data part ...
i have 3 environments ( /etc/puppet/environment/{development, stage,
production} ) ..
I am using gitolite for all my module repositories ... a
I have solved it as below with a wrapper to the zabbix module:
class common::zabbix {
contain zabbix::agent
Class['common::repos'] -> Class['common::zabbix']
}
and then in selected yaml environments
## YAML
---
classes:
- common
- ntp
- common::zabbix
El viernes, 17 de abril de 2
Hi There,
After a couple of years working with puppet I'm starting to use Hiera as
ENC with the hiera_include('classes').
Now I'm having problems with the resource ordering. I created a module to
set up my repos and then use some others modules. My problem is that my
'common' module must run b
On Friday, April 17, 2015 at 8:43:27 AM UTC-5, Ken Barber wrote:
>
> > Unfortunately, this only works for resource types that have an 'ensure'
> > parameter, which Yumrepo does not.
>
> I was under the impression yumrepo was ensureable now:
>
> https://github.com/puppetlabs/puppet/blob/master/
On Friday, April 17, 2015 at 3:25:37 AM UTC-5, jamese wrote:
>
> I've run into the same issue before. The way I resolved it was to add a
> file resource for each yum repo, just ensuring it is present. That way,
> the recursive purge on /etc/yum.repos.d/ doesn't nuke the repo files that
> pup
> Unfortunately, this only works for resource types that have an 'ensure'
> parameter, which Yumrepo does not.
I was under the impression yumrepo was ensureable now:
https://github.com/puppetlabs/puppet/blob/master/lib/puppet/type/yumrepo.rb#L16-L17
and https://docs.puppetlabs.com/references/lates
On Friday, April 17, 2015 at 3:41:03 AM UTC-5, Gavin Williams wrote:
>
> Chris
>
> Take a look at the 'resources'[1] type, which you can use to purge
> unmanaged resources of a specific type.
>
> So rather than purging the entire directory using a file resource, you
> just purge the yum-repo
I forgot to include this in my last post. We created a defined type to
wrap the yumrepo resource type. The defined type then declares the yumrepo
resource and the required file resource.
You could do the same, passing your defined type instead of yumrepo to the
create_resources function.
The
Thank you for the replies,
@Gavin, I attempted as you suggested but it still suffers from amnesia
regarding the repos themselves. But it did get me looking at other work
arounds and reading through to see if I can find a way for puppet with a
meta parameter or such to recognize these files exi
And the
link: https://docs.puppetlabs.com/references/latest/type.html#resources
Gav
On Friday, 17 April 2015 09:41:03 UTC+1, Gavin Williams wrote:
>
> Chris
>
> Take a look at the 'resources'[1] type, which you can use to purge
> unmanaged resources of a specific type.
>
> So rather than purg
Chris
Take a look at the 'resources'[1] type, which you can use to purge
unmanaged resources of a specific type.
So rather than purging the entire directory using a file resource, you just
purge the yum-repo files themselves...
Something like:
resources {'yumrepo':
purge => true
}
should
I've run into the same issue before. The way I resolved it was to add a
file resource for each yum repo, just ensuring it is present. That way,
the recursive purge on /etc/yum.repos.d/ doesn't nuke the repo files that
puppet has created. Arguably the yumrepo type should do that but it
doesn'
On Apr 16, 2015, at 11:40 PM, Michael Stahnke wrote:
> In particular, isn’t Puppet vulnerable to this problem?
> https://www.ruby-lang.org/en/news/2015/04/13/ruby-openssl-hostname-matching-vulnerability/
>
> Only if you're using a 3rd party CA , which 99.9% of users do not do and
> using wildca
21 matches
Mail list logo
