Hi
> class users {
>
> user { "ben" :
> ensure => present,
> uid => 1010,
> gid => "users",
> managehome => true,
> password => "blahblah"
> }
>
> ssh_authorized_key { "ben-key":
>ensure => present,
>key=> "blahblah",
>type => "ssh-dss",
>user => "be
Hi. I'm running facter 1.5.1 in Ubuntu Jaunty. facter is giving me
inconsistent results for the operatingsystem fact:
$ facter operatingsystem
Debian
$ facter | grep operatingsystem
operatingsystem => Ubuntu
operatingsystemrelease => 9.04
Is this a known problem? Possibly fixed in a more rec
Hello. I am configuring my puppetmaster with the authorized_keys
management recipe from the recipes repository on the Puppet site. So
far, it does what it should *except* generate a keypair for new users,
and I'm unsure of how to proceed from here. Any suggestions would be
greatly appreciated.
I
Ben Lavender schrieb:
> Hi,
>
> I'm trying to create an authorized key and a user in one go and I'm
> not getting anywhere:
>
> class users {
>
> user { "ben" :
> ensure => present,
> uid => 1010,
> gid => "users",
> managehome => true,
> password => "blahblah"
> }
>
> ssh
> Which seems odd because the user's account and home directory do
> exist. In this example, the puppet client and puppetmaster are the
> same machine. I can reproduce the problem with a second machine as the
> client, as well.
do exists, means that you have an actual file-resource managing it? I
> Are you using 0.25? Then you might also hit a current bug where traling
> slashes are stripped of but not in relationships. So you might want to
> change File["${homeroot}/$name/"], to File["${homeroot}/$name"],
ftr: http://projects.reductivelabs.com/issues/2675
cheers pete
--~--~-~--
On Nov 12, 11:51 am, Ben Lavender wrote:
>
> > you have to specify relationship between these 2 resources. This means
> > that the ssh_authorized_key needs a require => User['ben']
>
> This is not the issue; specifying the relationship manually does not
> solve the problem. Further, the docs sta
Hi all,
I would not like to confuse people.
What I described in OT only happens to pakiti service and not to yum.
So I suppose it's a "service" specific problem. It has its own init
script but must behaves diff from standrd RH service.
Cheers,
Arnau
--~--~-~--~~~-
Arnau Bria wrote:
> # /etc/init.d/pakiti status
> nightly pakiti update is enabled
> # /etc/init.d/pakiti stop
> Disabling pakiti: [ OK ]
> # /etc/init.d/pakiti status
> nightly pakiti update is disabled
Now try this:
# /etc/init.d/pakiti start
Dear all,
I want to manage services in my AIX node. I mean i can start/stop
service using
exec { command => "/etc/rc/blah-service start",} but i like this to be
done by puppet itself with its type reference 'service'.
has anyone manage the services on AIX machine ??? if yes, please share it!
P
jcbollinger wrote:
> On Nov 10, 10:13 am, Thomas Bellman wrote:
>> The problem is that 'chkconfig on' does an implicit add of
>> the service; but it does a half-assed job, in that it only adds the
>> start links, not the kill links. Thus, it is very easy to get into
>> the broken state by doin
On Nov 11, 12:18 pm, David Schmitt wrote:
> As a debianista I obviously haven't touched chkconfig ever. But in the
> spirit of DWIM, I would expect a provider to do the magic to get (in
> this case) an init-script under the control of the chosen tool.
I'd be happier with that plan if the WIM we
On Nov 11, 10:13 am, Arnau Bria wrote:
> I'm just trying to control a service that does not run a process but
> enables a cron (like yum autoupdate or pakiti).
What Bellman said. Puppet is limited in what it can do for you if
your initscipt's status command doesn't return an accurate status.
Hi Thomas,
that's the problem
all status return code is 0...
Thanks, going to contact developer and see if he accepts a patch.
Cheers,
Arnau
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Puppet Users" grou
hi all,
I am trying to create a set of users on my AIX node and trying to get
their authorized_keys which are already hosted on my server with name
like, myuser_id_ds.pub. Currently i am managing 2 nodes (1. Suse
Enterprise 2. AIX). I defined the 'source' file paths in 2 separate
contexts in
FYI: for those looking for examples and more discussions.
I've put my collection of infrastructure testing links online:
http://www.jedi.be/blog/2009/11/12/collection-of-test-driven-infrastructure-links/
--~--~-~--~~~---~--~~
You received this message because yo
We treat zones like normal nodes for the most part. There is a
separate base class that is zone specific but otherwise we don't do
things any differently.
-- cwebber
On Nov 12, 2009, at 8:28 AM, windowsrefund wrote:
>
> Getting past zone creation, I'm curious to know how people are
> managi
Getting past zone creation, I'm curious to know how people are
managing zone-specifics like users, packages, and services.
Looking forward to this thread...
Adam
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Pupp
On Nov 12, 11:32 am, Christopher Webber wrote:
> We treat zones like normal nodes for the most part. There is a
> separate base class that is zone specific but otherwise we don't do
> things any differently.
>
So each zone runs a puppet client?
--~--~-~--~~~---~
On Nov 12, 11:32 am, Christopher Webber wrote:
> We treat zones like normal nodes for the most part. There is a
> separate base class that is zone specific but otherwise we don't do
> things any differently.
>
So each zone runs a puppet client?
--~--~-~--~~~---~
Yes.
-- cwebber
On Nov 12, 2009, at 8:55 AM, windowsrefund wrote:
>
>
>
> On Nov 12, 11:32 am, Christopher Webber wrote:
>> We treat zones like normal nodes for the most part. There is a
>> separate base class that is zone specific but otherwise we don't do
>> things any differently.
>>
>
> So
That's not an option for me as I'm already planning on managing a few
hundred physical nodes. Any other approaches out there?
Thanks,
Adam
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To post
On 11/12/09 12:10 PM, "windowsrefund" wrote:
> That's not an option for me as I'm already planning on managing a few
> hundred physical nodes. Any other approaches out there?
We manage a few thousand. We generally operate on "groups" (as defined in
our external nodes classifier) and never on i
Brian,
Can you be more specific? For example, how would you go about ensuring
that apache is running inside a non-global zone?
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To post to this gr
On 11/12/09 1:36 PM, "windowsrefund" wrote:
> Can you be more specific? For example, how would you go about ensuring
> that apache is running inside a non-global zone?
The same as we do on any node. Service ensure running.
--
Brian Akins
--~--~-~--~~~---~--~~
On Nov 12, 2:35 am, Peter Meier wrote:
> > Which seems odd because the user's account and home directory do
> > exist. In this example, the puppet client and puppetmaster are the
> > same machine. I can reproduce the problem with a second machine as the
> > client, as well.
>
> do exists, means t
Brian,
OK so you're also running a puppet client on each virtual node.
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To post to this group, send email to puppet-users@googlegroups.com
To unsu
Andrew Schulman writes:
> Hi. I'm running facter 1.5.1 in Ubuntu Jaunty. facter is giving me
> inconsistent results for the operatingsystem fact:
> $ facter operatingsystem
> Debian
> $ facter | grep operatingsystem
> operatingsystem => Ubuntu
> operatingsystemrelease => 9.04
> Is this a kn
On Thu, Nov 12, 2009 at 11:29 AM, Russ Allbery wrote:
>
> Andrew Schulman writes:
>
> > Hi. I'm running facter 1.5.1 in Ubuntu Jaunty. facter is giving me
> > inconsistent results for the operatingsystem fact:
>
> > $ facter operatingsystem
> > Debian
>
> > $ facter | grep operatingsystem
> >
Nigel Kersten writes:
> Ugh. This is particularly sucky, and I vote this fact needs to be fixed
> for this irrespective of the plugin loading issue.
> The problem is:
> if Facter.value(:lsbdistid) == "Ubuntu"
>"Ubuntu"
> elsif FileTest.exists?("/etc/debian_version")
On 10/29, Peter Meier wrote:
> > I'm trying to have a exec dependency on a service object that would keep
> > it from being restarted if the exec fails. Unfortunately, the service
> > gets refreshed regardless whenever the exec is run, failure or not.
[...]
>
> don't set the command to refreshon
On Nov 12, 7:32 am, M F Haris wrote:
> I want to manage services in my AIX node. I mean i can start/stop
> service using
> exec { command => "/etc/rc/blah-service start",} but i like this to be
> done by puppet itself with its type reference 'service'.
>
> has anyone manage the services on AIX
I can get behind that. Nominated :)
On Nov 12, 2009, at 12:52 AM, Michael T. Halligan wrote:
>
> I just nominated Reductive Labs for a Crunchie in the "Best Enterprise
> Startup" category, if only because it's the kind of ridiculous SF startup
> popularity game that I think would annoy Luke
I would like to set up two puppetmaster servers, one slaving off the
other. Can anyone point me towards some documentation for setting it
up? I've looked, but haven't had any luck finding anything.
--~--~-~--~~~---~--~~
You received this message because you are s
Here are the two go-to points for me:
http://reductivelabs.com/trac/puppet/wiki/ModuleOrganisation
http://www.devco.net/archives/2009/09/28/simple_puppet_module_structure.php
-=Eric
On Nov 10, 2009, at 4:38 PM, Allan Marcus wrote:
>
> CAn someone point me to some documentation on the autoloa
I'm looking for people and organisations I can work with in the London
area who can provide tangible problems, alpha test and give feedback
on puppet win32 support in the forthcoming months.
Ideallly you have a heterogenous setup and some specific problems to
solve, and potentially happy to have
Hey All,
I'm wondering if there is an easy way to distinguish between SP2 and SP3 on a
SLES 10 installation.
Thanks,
Matt Delves
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To post to thi
Is anyone using multiple environments? Can someone at least tell me
what they did to get it working?
Thanks,
Scott
On Nov 12, 6:53 am, Scott wrote:
> So I'm trying to get multiple environments to work with puppet 0.25.1
> on ubuntu 8.04 and no matter what I do, puppet just completely ignores
>
I just had to set the environment variable in the puppet.conf file before
starting the puppetd daemon. Then in each class I use that variable to
determine what happens, for example
file { "smtpd.conf":
owner => root,
group => root,
mode=> 600,
On Thu, Nov 12, 2009 at 2:36 PM, Russ Allbery wrote:
>
> Nigel Kersten writes:
>
> > Ugh. This is particularly sucky, and I vote this fact needs to be fixed
> > for this irrespective of the plugin loading issue.
>
> > The problem is:
>
> > if Facter.value(:lsbdistid) == "Ubuntu"
> >
It might be in /etc/suse-release if Facter isn't already picking it up
On Nov 12, 2009, at 2:27 PM, Matt Delves wrote:
>
> Hey All,
> I'm wondering if there is an easy way to distinguish between SP2 and SP3 on a
> SLES 10 installation.
>
> Thanks,
> Matt Delves
>
>
> >
--~--~-~--
Scott,
Multiple environments is working for me under puppet 0.25.1.
I'm not sure what's wrong, but I wonder if having a manifest and
modulepath entry in the [main] part of your puppet.conf is overriding
any sections below. My puppet.conf does not have a manifest or
modulepath entry in [main].
P
Here's the solution I wound up using. I created a puppet.conf.erb
template file and had puppet create the puppet.conf file dynamically.
The critical pieces (I only wanted directories starting with v, like
v01, v02, et cetera):
<%
dirs = Dir.entries("/etc/puppet/manifests").sort
-%>
[main]
...
[
On Thu, Nov 12, 2009 at 4:48 PM, lance dillon wrote:
>
>
> On Thu, Nov 12, 2009 at 2:36 PM, Russ Allbery wrote:
>
>>
>> Nigel Kersten writes:
>>
>> > Ugh. This is particularly sucky, and I vote this fact needs to be fixed
>> > for this irrespective of the plugin loading issue.
>>
>> > The probl
Pete,
I only put the "modulepath" and "manifest" entries in the "main"
section in an attempt to get environments to work.
So your configuration is working with just a separate section like
[testing] in my puppet.conf? No "environments" entry in
[puppetmasterd] and just "modulepath" and "manifes
> This has nothing to do with grep.
>
> facter should report a single consistent value for 'operatingsystem' when
> invoked in either of these ways.
Agreed.
> Andrew, have you put a bug report in about this?
No, I wanted to check first if I was missing something. I'll file a
bug report about i
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Andrew Schulman wrote:
>> This has nothing to do with grep.
>>
>> facter should report a single consistent value for 'operatingsystem' when
>> invoked in either of these ways.
>
Maybe I am jumping in too late but do you have lsb installed on that
ho
James Turnbull writes:
> Andrew Schulman wrote:
>>> This has nothing to do with grep.
>>> facter should report a single consistent value for 'operatingsystem' when
>>> invoked in either of these ways.
> Maybe I am jumping in too late but do you have lsb installed on that
> host (lsb-base packa
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Russ Allbery wrote:
>
> Yeah, I bet that's more likely than my explanation, given that it's also
> not working inside Puppet (which doesn't have the plugin problem).
> Although that doesn't explain why running facter with grep works
>
Ah missed
49 matches
Mail list logo
