If you are actually managing apt and need to trigger an apt-update
before a package install I can highly recommend the puppetlabs/apt
module on puppet forge I am using it in my environments and it does a
very good job at managing all of apt and friends.
https://forge.puppetlabs.com/puppetlabs/apt
The ENC can pass on any variable you wish to define and they can be used by
Hiera.
I haven't got around to implementing a true CMDB yet so I use a very simple
ruby script as my ENC. It reads the hieradata/nodes/fqdn.yaml file for a
"role" value and returns that in the output to puppet. My hiera.yam
Hey
Thanks for the reply
I want the ENC to pass on variables or the like so that these can be used
by hiera to apply roles/profiles etc.
eg
ENC returns a nodes software role eg role: desktop_software
Then hiera can load the desktop_software.yaml file to apply that software
role as part of i
On 07/24/14 17:17, Supriya Uppalapati wrote:
>
> I did like this using file_line resource
> file_line { 'replace a line to /home/infauser/.bash_profile':
> path => '/home/infauser/.bash_profile',
> line => 'export INFA_CODEPAGENAME=UTF-8',
> #match => '^# export INFA_CODEPAGENAM
For example, I have an array like this:
clusters => [
'cluster1',
'cluster2',
'cluster3',
'cluster4',
]
I then have a cluster.init.erb that looks soemthing like this:
start on runlevel [2345]
stop on runlevel [!2345]
expect fork
respawn
<% @clusters.each do | cluster | -%>
env PIDF
Hi CD,
Just tried these too, but somehow puppet is not honoring any variables in
the resources.
baseurl => http://192.168.1.100/centos/$releasever/os/$basearch,
and
baseurl => "http://192.168.1.100/centos/$releasever/os/$basearch";,
Somehow, notify is working fine.
notify {'FYI': mess
Hi Vikas,
This is not the solution to your question but possibly alternative option.
It seems you want to generate a yum repo url for specific OS version so
that it picks-up correct version arch etc.
How about if you use following and set up repos to match.
baseurl=http:///centos/$releasever/o
Adding this for informational purposes.
This problem was related to the one I posted in another thread. The fix is
below.
Blank the file.
# cp /dev/null /var/lib/puppet/state/state.yaml
Run the agent.
# /opt/puppet/bin/puppet agent --test
notice: /Stage[main]/Solaris/File[/etc/logadm.conf]/ensu
Thanks for the pointer. I believe this does it.
No need to stop the puppet agent. Just two things:
- blank the /var/lib/puppet/state/state.yaml file
- run the puppet agent
New values are recorded.
Blank the file.
# cp /dev/null state.yaml
Run the agent.
# /opt/puppet/bin/puppet agent --test
n
Dear friends,
I wonder whether anyone have any experience deploying a Linux HA cluster
using Puppet. I would like to automate the deployment of HA cluster
(http://clusterlabs.org/doc/en-US/Pacemaker/1.1-pcs/html/Clusters_from_Scratch/index.html)
in a Puppet master-agent model.
These are the c
Hi Sean/Juan,
Yes, the redhat-lsb package is installed on my RedHat 6.4 machine which is
my Puppet Client server. Puppet server is Ubuntu 14.04.
root@box27:~# rpm -qa | grep redhat-lsb
redhat-lsb-graphics-4.0-7.el6.x86_64
redhat-lsb-compat-4.0-7.el6.x86_64
redhat-lsb-core-4.0-7.el6.x86_64
redhat
I presume you just want the information for the one customer, and not all
customers?
If the first, I assume you set the customer name, and you just dump the
information in a file named after the customer. If you want all customers,
dump it all in one common file.
Whilst, we're still using extlook
I don't believe that is possible.
You will probably need to collate all the files in customers/ into one file.
I am going to hazard a guess that you have data relating to your
customers in that directory.
If the data in those files is specific to each customer and each of
your nodes had only one c
On 24 Jul 2014 21:50, "Peiven Cheng" wrote:
>
> Hi all,
>
> We use mcollective to trigger puppet sync in windows servers.
> While we send the 'mco puppet runonce' to trigger 8~10 servers doing puppet
> agent apply,
> it occasionally happened the error of
> Could not retrieve catalog from remote s
On Thu, Jul 24, 2014 at 1:22 PM, huhm4n wrote:
> Thanks a lot for the idea. I'll be creating a facter and see how it works
> or what i can do with it
>
>
> On Thursday, July 24, 2014 8:25:21 AM UTC-7, showy wrote:
>
>> Hi, for this problem you would be better off with other tools.
>>
>> Neverthe
Thanks a lot for the idea. I'll be creating a facter and see how it works
or what i can do with it
On Thursday, July 24, 2014 8:25:21 AM UTC-7, showy wrote:
>
> Hi, for this problem you would be better off with other tools.
>
> Nevertheless you could create a facter thar list all the users in th
Hi,
I have two ceph clusters 'ceph_prod_cluster' and 'OSC_acc_storage'
For the OSC_acc_storage cluster I have 3 ceph monitor nodes
isp-oscaccstormon-01, isp-oscaccstormon-02 and isp-oscaccstormon-03. I
deploy the nodes with puppet-ceph. Like you can see on
https://github.com/mpiscaer/puppet-ceph
I was vaguely nosy, and found a note that the checksum is stored in state.yaml.
Maybe stop puppet, nuke the state file, and start puppet?
http://puppetlabs.com/blog/all-about-auditing-with-puppet
/var/lib/puppet/state/state.yaml
NB: Untested advice, potentially dangerous, use at your own risk.
Does anyone have a tip regarding this?
How do I stop it?
I know that the "audit=>all" is causing this.
It recorded the MD5 sum of the file when the audit ran for the first time.
But a newer file was created as a result of patching the client machine.
Is there a command I could rerun on the client
Hi,
You can try to execute facter in the nodes, with this you can see the
facter availables.
It's posibble that *lsbdistrelease *is not available, but
*operatingsystemrelease* works properly.
This is because *lsbdistrelease *uses a package and *operatingsystemrelease*
read from a file /etc/re
Hi all,
I have a question for the experts in the group:
say there's an exec resource dependent on some other resource and
refreshonly=> true. As we know already, a change in the dependent resource
will send a refresh event to the exec resource. This will however be
scheduled for AFTER the chan
Grant,
You probably want to use an ENC, which will allow you to define any fact
from a specified datasource.
In a simple implementation I use an ENC to read the node's hiera file and
output a few facts, such as a node's role/group, which are then used by
hiera to define additional information for
Hi, for this problem you would be better off with other tools.
Nevertheless you could create a facter thar list all the users in the
system and use that information in a class to check and save against a
centralized db.
On Jul 23, 2014 4:35 PM, "huhm4n" wrote:
> I'm trying to audit local users
Hi,
I did like this using file_line resource
*file_line { 'replace a line to /home/infauser/.bash_profile': path =>
'/home/infauser/.bash_profile', line => 'export
INFA_CODEPAGENAME=UTF-8', #match => '^# export
INFA_CODEPAGENAME=UTF-8.*',*
*But it is giving me an error:*
On 2014-24-07 14:27, Werner Flamme wrote:
em...@benjaminmertens.de [24.07.2014 12:42]:
# site.pp
define iterateSystemRole {
case $name {
'app1': {
include app1
$defineVariable = 'content'
}
On Wednesday, July 23, 2014 6:26:00 PM UTC-5, Stack Kororā wrote:
>
> Greetings!
>
> Thank you so much John. I just learned something new about Puppet. Utilizing
> inline_template is a heck of a lot easier then how I first attempted that
> variable substitution. I might have to go back and fix
Thanks Hunter - this is the info I was looking for. Now of course I need
puppet to create this file for me since the idea here is that I dont have
to touch the node. Can you be more specific on he "just add your own
info" Is the the client/socket/unset info? can you populate those with
d
Hi,
I would like to know if exists the possibility to include all the yaml
files of a directory in hiera.
The hiera config looks like:
:hierarchy:
- %{hostname}
- secrets
- customers/*
:backends:
- yaml
:yaml:
:datadir: '/etc/puppet/data'
Cheers,
Juan Moreno
--
You received this m
em...@benjaminmertens.de [24.07.2014 12:42]:
>
> # site.pp
>
> define iterateSystemRole {
> case $name {
> 'app1': {
> include app1
> $defineVariable = 'content'
> }
> 'app2': {
>
Hi Vikas,
Do you have the package 'redhat-lsb' installed? The $lsbdistrelease fact is
only populated when this package is installed
Sean
On 24 July 2014 17:07, Vikas Kumar wrote:
> Hello All,
>
> I have a kickstart server which I use to deploy many minor versions of
> CentOS/RHEL 5 and 6 for
Hello All,
I have a kickstart server which I use to deploy many minor versions of
CentOS/RHEL 5 and 6 for both 32 and 64 Bit OS.
I have lots of yum repository files. For. e.g. - RHEL_5.5_64.repo,
RHEL_5.6_64.repo, RHEL_6.2_32, RHEL_6.5_64, etc. I am looking to use Puppet
to automatically deplo
Hi,
my servers have some kind of "role" like "app1" or "app1,app2" which im
getting out of a selfmade global facter variable.
Im then going through this list and include classes depending on the
content of this variable.
# site.pp
define iterateSystemRole {
case $name {
Hi all,
We use mcollective to trigger puppet sync in windows servers.
While we send the 'mco puppet runonce' to trigger 8~10 servers doing puppet
agent apply,
it occasionally happened the error of
Could not retrieve catalog from remote server: Error 400 on SERVER: Could
not find class test_1_0_
I am looking to implement puppet coming from cfe2
I am also finding it hard to make the connection between nodes and groups
using hiera. I would like a solution where I can define host groups etc on
the puppet master rather than relying on the client to have a file(s) put
in particular location
34 matches
Mail list logo
